update article

docs/en/Community-Articles/2026-06-25-Convex-Summit-2026-Recap/Post-v2.md

@@ -1,207 +0,0 @@
-### 1. Alternative Article Title Suggestions
-
-1. **Unifying Dev, Architecture, and AI: My Experience Speaking on Conversational SQL at CONVEX 2026**
-2. **CONVEX Summit 2026: Notes from the Stage, the Cinema Screens, and the Future of Database-to-Agent Systems**
-3. **Beyond the Vibe Coding: Speaking at CONVEX 2026 and Re-Architecting Conversational B2B Systems**
-
-### 2. Selected Title
-
-# Unifying Dev, Architecture, and AI: My Experience Speaking on Conversational SQL at CONVEX 2026
-
-### 3. Meta Description
-
-Join Volosoft Co-Founder Alper Ebiçoğlu as he shares his firsthand experience speaking at CONVEX Summit 2026 in Madrid, exploring natural-language-to-SQL architecture, Model Context Protocol (MCP), and lessons in AI security.
-
-### 4. SEO Slug Suggestion
-
-```
-convex-summit-2026-speaks-view-conversational-sql
-```
-
-### 5. Full Article
-
-Arriving in Madrid this June for the inaugural CONVEX Summit 2026 felt like witnessing a major shift in how our industry talks about building software. For years, developers, database administrators, and software architects have operated in distinct technical silos. We have watched artificial intelligence disrupt daily operations, development teams rapidly adopt new frameworks, and distributed architectures grow increasingly complex. Yet, these massive shifts have largely occurred in parallel.
-
-To celebrate their twentieth anniversary, the team at Plain Concepts took a bold, necessary step: they unified three of Spain’s flagship tech events—dotNET, the Global Software Architecture Summit (GSAS), and Singularity Tech Day—into a single, cohesive experience.
-
-As I walked into the Kinépolis Ciudad de la Imagen—the largest cinema complex in Madrid—the scale of this integration was immediately apparent. The venue was buzzing with over 1,200 tech professionals representing more than 25 countries. For me, as a co-founder and software architect at Volosoft, this was more than just another conference. It was a unique convergence point where theory met execution, and where I had the privilege of taking the main stage to speak on a topic I’ve been living and breathing: re-architecting how enterprise applications talk to databases.
-
-!(convex_keynote_stage.jpg)
-
-## Speaking at CONVEX: Chat with Your Data
-
-My session, titled **"Chat with Your Data: Turn any database into a conversational reporting engine,"** was scheduled in front of an incredibly engaged audience of developers, CTOs, and systems architects. The primary problem I wanted to tackle is one that almost every B2B application team faces: the endless cycle of custom report building. Traditional enterprise applications are bottlenecked by the constant demand for custom queries, Excel exports, and visual dashboards.
-
-My talk introduced a conversational reporting approach designed to let non-technical stakeholders safely generate complex reports from their database simply by chatting—as if they were messaging a human analyst.
-
-!(alper_ebicoglu_presentation.jpg)
-
-On stage, I detailed the exact architectural pipeline we built using.NET to securely connect natural language prompts to database engines. Letting an LLM generate SQL queries is easy in a demo, but incredibly dangerous in an enterprise environment. If you simply pass user input directly to an LLM and run the resulting string against your database, you are inviting disastrous SQL injections, massive context bloat, and uncontrolled resource exhaustion.
-
-To solve this, we designed a multi-stage validation pipeline that prioritizes security and performance:
-
-```
-[User Natural Language Input]
-              │
-              ▼
-   ───► Minimize schema metadata injected into prompt
-              │
-              ▼
-    ──► Parse user intent to avoid context bloat
-              │
-              ▼
-     ──────► Draft query based on precise prompt constraints
-              │
-              ▼
-    ─► AST parsing to block DDL/DML, enforce read-only
-              │
-              ▼
-   ────► Safe execution isolated from production database
-              │
-              ▼
-  [Output Generation Engine] ───► Dynamic formatting into Excel sheets & charts
-```
-
-### The Security Mathematics of SQL Validation
-
-The most critical phase of this pipeline is our dynamic SQL parser. Before any generated query touches the database, the.NET application parses the string into an Abstract Syntax Tree (AST). This allows us to run a deterministic evaluation of the query structure.
-
-We can model this strict security boundary mathematically. Let $Q$ be the generated SQL query, and let $T(Q)$ be the set of operation tokens identified in the AST. We define the security function $S(Q)$ as:
-
-$$S(Q) = \begin{cases}  1 & \text{if } T(Q) \subseteq \{\text{SELECT}\} \land T(Q) \cap \{\text{INSERT}, \text{UPDATE}, \text{DELETE}, \text{DROP}, \text{ALTER}, \text{CREATE}\} = \emptyset \\  0 & \text{otherwise}  \end{cases}$$
-
-If $S(Q) = 0$, the query is immediately rejected at the application boundary, completely mitigating malicious prompt injections or model hallucinations before they can cause damage. We also run these validated queries exclusively against an isolated read-replica, completely separating conversational reporting workloads from our primary transaction database.
-
-During the session, I demonstrated how the pipeline extracts metadata to construct the schema map, routes user intents, and dynamically compiles the resulting database rows into structured Excel files and interactive charts. It was highly rewarding to hear from the community afterward about how this approach solves real-world security concerns while dramatically improving the B2B developer experience.
-
-## What I Learned From the English Sessions
-
-When I wasn't on stage or talking with attendees, I spent my time attending the English-language sessions. Because the conference unified dotNET, GSAS, and Singularity, the technical depth across the tracks was remarkable. Several sessions provided profound, second-order insights into how enterprise engineering teams are actually putting AI and advanced architecture patterns to work at scale.
-
-### AI, Security, and System Exploits
-
-Chema Alonso's Keynote on Day 2, *"Hacking ( with | the ) AI,"* was an eye-opening deep dive into the darker side of generative systems. Alonso, a leading security figure, showcased how malicious actors are actively utilizing AI to accelerate the development of system exploits.
-
-What struck me most was his analysis of semantic vulnerabilities. Traditional firewalls and security protocols are completely blind to threat vectors like jailbreaking, prompt injection, and model exfiltration. Alonso's core thesis resonated deeply with my own presentation: AI is a powerful assistant, but it cannot be treated as a security boundary. If you build an AI feature, you must assume the output generated by the model is untrusted and validate it with rigorous, deterministic code.
-
-### Redefining the Next Digital Frontier with MCP
-
-Another highly practical session was delivered by Manuel Sanchez and Carlos Mendible, titled *"AI, Agents and MCP: Redefining the Next Digital Frontier"*. They introduced the Model Context Protocol (MCP)—an emerging open standard designed to structure how AI agents interact with local applications, databases, and development environments.
-
-Sanchez and Mendible highlighted a common mistake developers make when building agentic integrations: exposing raw CRUD (Create, Read, Update, Delete) database tables to the model's context window. This "context bloat" dramatically increases token costs and degrades the agent's reasoning speed.
-
-Instead, they demonstrated how MCP servers should expose high-level, parameter-driven business tools (e.g., executing a specific calculation or pulling a pre-filtered report). This approach pushes computation back onto the database or backend systems, saving tokens and keeping agents highly performant.
-
-| **Integration Pattern** | **Context Bloat (Tokens)**                          | **Latency**                                    | **Security Control**                        |
-| ----------------------- | --------------------------------------------------- | ---------------------------------------------- | ------------------------------------------- |
-| **Raw CRUD Exposure**   | Extremely High (Exposes raw schema & raw tables)    | High (Model must process entire dataset)       | Very Poor (Relying on model constraints)    |
-| **MCP Business Tools**  | Minimal (Exposes targeted APIs/parameterized tools) | Low (Database/backend handles heavy computing) | Excellent (Enforces strict code boundaries) |
-
-### Re-Evaluating Architectural Trade-offs and Climate Impact
-
-Eoin Woods, one of the leading figures in software design, brought invaluable perspective to the GSAS track with his talk, *"The Key to the Prisoners' Dilemma"*. Woods discussed the constant tug-of-war between business speed and long-term architectural stability, using game theory to prove that proper architecture is actually the primary vehicle for sustainable, ongoing business value.
-
-What made Woods' contribution even more fascinating was his work on green software engineering. He pointed out that the carbon emissions of global computing infrastructure are rising rapidly, with ICT emissions projected to reach 5% by 2030—driven in large part by the extreme computational demands of modern AI models.
-
-Woods introduced the concept of **Demand Shifting**. By utilizing smart, orchestrating software architectures, enterprise teams can dynamically route heavy, non-time-sensitive AI training and query workloads to data centers currently operating on clean, excess renewable energy. This simple architectural decision can reduce operational emissions by up to 40% with virtually zero impact on system performance.
-
-## The Conference Experience
-
-Kinépolis Madrid proved to be an outstanding venue for a tech summit of this scale. Showing complex architecture diagrams, SQL configurations, and C# code on IMAX-sized movie screens was a developer's dream. The audio clarity and amphitheater seating ensured that even the most dense, code-heavy presentations felt incredibly engaging.
-
-!(convex_networking_hall.jpg)
-
-But beyond the high-quality presentation rooms, what really set CONVEX apart was the lack of superficial commercial noise. There were no aggressive sales pitches or standard marketing booths trying to reel you in. Instead, the networking areas were filled with genuine technical conversations.
-
-During the coffee breaks and lunch sessions, I spent hours talking with developer advocates, CTOs, and software architects representing the international.NET and open-source communities. We compared notes on our experiences with Blazor WebAssembly, discussed scaling multi-tenant SaaS structures, and debated the practical limits of "vibe coding". The community-first energy was palpable, showing that the real value of these events is built on the shared experiences and connections made off-stage.
-
-## My Key Takeaways
-
-Reflecting on my conversations, my presentation, and the excellent sessions I attended, several core takeaways stand out for any B2B engineering leader:
-
-- **AI features require absolute data boundaries:** Building conversational database features is a powerful way to eliminate custom report backlogs, but you must validate LLM-generated outputs before they reach your data. Never allow an AI model to write directly to a production database, and always validate queries using AST analysis.
-- **The Model Context Protocol (MCP) is the new standard:** Rather than creating custom, ad-hoc integrations for every agent, we must design modular MCP servers that expose clean, business-level APIs to AI models. This reduces context bloat and enforces a cleaner separation of concerns.
-- **Green software is an architectural priority:** With AI dramatically increasing energy consumption, we can no longer ignore the environmental impact of our software systems. Implementing patterns like Demand Shifting to run heavy workloads during green energy peaks is becoming a vital non-functional requirement.
-- **Developer experience remains a massive competitive advantage:** The success of tools like the ABP Framework and pre-built modular architectures is proof that B2B development teams want to focus on business logic rather than writing repetitive boilerplate code. By automating routine tasks like query generation and report compilation, we can free up engineering teams to focus on core platform value.
-
-## Closing
-
-The inaugural CONVEX Summit 2026 was a resounding success. Plain Concepts did an incredible job of transforming three separate industry dialogues into a unified, high-impact event that reflected the real challenges tech organizations face today.
-
-I want to extend my sincere thanks to the organizers, especially Ivan Suárez Álvarez, for putting together such a high-caliber event. I'm also deeply grateful to all the speakers who shared their hard-earned production lessons, and to every member of the.NET and Volosoft communities who stopped by to talk, share feedback, and celebrate our shared passion for building high-quality software.
-
-I left Madrid with a notepad full of new architectural ideas, a stronger network of global peers, and an even deeper conviction that the intersection of structured software architecture and generative AI is the most exciting place to be building right now. I cannot wait to see where these conversations take us, and I look forward to returning for the next edition!
-
-### 6. Used Photos and Placement Details
-
-1. **Görsel Dosya Adı**: `convex_keynote_stage.jpg`
-   - **Yerleştirilen Bölüm**: Kısa giriş (Introduction) bölümünün hemen altı.
-   - **Alt Metin (Alt Text)**: The grand keynote stage at Kinépolis Ciudad de la Imagen welcoming over 1,200 international technology leaders to CONVEX 2026.
-2. **Görsel Dosya Adı**: `alper_ebicoglu_presentation.jpg`
-   - **Yerleştirilen Bölüm**: "Speaking at CONVEX: Chat with Your Data" ana başlığının hemen altı.
-   - **Alt Metin (Alt Text)**: Alper Ebiçoğlu presenting 'Chat with Your Data' live on stage, detailing the pipeline that bridges natural language with secure enterprise database queries.
-3. **Görsel Dosya Adı**: `alper_slide_ast_validation.jpg`
-   - **Yerleştirilen Bölüm**: "Speaking at CONVEX: Chat with Your Data" bölümündeki teknik SQL analizi ve matematiksel formülün hemen altı.
-   - **Alt Metin (Alt Text)**: An architecture slide showing the schema discovery, LLM processing, and query validation pipeline of the conversational reporting engine.
-4. **Görsel Dosya Adı**: `convex_networking_hall.jpg`
-   - **Yerleştirilen Bölüm**: "The Conference Experience" bölümünün hemen altı.
-   - **Alt Metin (Alt Text)**: Attendees engaging in technical discussions and B2B networking during the breaks in the exhibition hall of Kinépolis Madrid.
-
-### 7. LinkedIn Post Suggestions
-
-#### Post 1: Short & B2B Professional (General Event Review)
-
-> Unifying dotNET, GSAS, and Singularity Tech Day, CONVEX Summit 2026 in Madrid brought together over 1,200 international tech leaders to answer a single question: How do we turn technological potential into real-world software impact?
->
-> I was thrilled to take the stage to talk about conversational database architectures. Read my complete B2B conference review for technical highlights on AI security, green computing, and agentic workflows: [Link] #CONVEX2026 #SoftwareArchitecture #EnterpriseAI #DotNet
-
-#### Post 2: Short & Technical (NL-to-SQL Pipeline)
-
-> Letting an LLM generate SQL queries is easy in a demo, but incredibly risky in production. At CONVEX 2026, I shared our.NET-based pipeline for "Chat with Your Data," demonstrating how to use Abstract Syntax Tree (AST) parsing to enforce strict read-only queries at the application boundary.
->
-> Curious about dynamic schema discovery, context injection, and preventing context bloat? Check out my latest technical write-up from the Madrid stage: [Link] #ConversationalSQL #SoftwareEngineering #PostgreSQL #B2BTech
-
-#### Post 3: Medium & Analytical (Architectural Focus)
-
-> AI-Guards alone will not save your B2B application. At CONVEX 2026, experts like Chema Alonso reminded us that generative systems are not security boundaries.
->
-> As software architects, we must assume LLM outputs are untrusted. In my latest article, I evaluate key architectural takeaways from the Madrid summit—exploring why we must transition to specialized Model Context Protocol (MCP) servers, why we should adopt "Demand Shifting" to lower the carbon footprint of heavy AI workloads, and how to safely design natural-language-to-SQL engines.
->
-> Read the full technical breakdown: [Link] #EnterpriseAI #CyberSecurity #SystemDesign #MCP
-
-#### Post 4: Medium & Developer Productivity (SaaS & Frameworks)
-
-> Developer experience remains the ultimate competitive edge. As creators of the open-source ABP Framework, we at Volosoft are always looking for ways to cut out boilerplate code and accelerate feature delivery.
->
-> At CONVEX 2026, the discussion shifted from "vibe coding" back to spec-driven architecture. By building conversational reporting tools that handle the data translations while our C# code handles validation and dynamic Excel generation, we can eliminate reporting bottlenecks forever.
->
-> Here are my reflections on how modern development, architecture, and AI are finally merging into a single, high-impact narrative: [Link] #DX #DotNet #SaaS #ABPFramework
-
-#### Post 5: Personal & Story-Driven (My Speaker Journey)
-
-> What an incredible week in Madrid! Speaking at the inaugural CONVEX Summit 2026 was an absolute highlight of my year. Sharing the stage at the stunning Kinépolis cinema venue to present our "Chat with Your Data" conversational reporting pipeline was a fantastic experience.
->
-> Beyond presenting, what made this trip truly special was the community. Meeting with fellow software architects at the speaker dinner, exploring historical tech challenges, and exchanging notes on modern.NET configurations over coffee made for some unforgettable conversations.
->
-> I want to extend a huge thank you to Plain Concepts and Ivan Suárez Álvarez for organizing a stellar event. I've gathered my favorite technical sessions, personal notes, and major architecture takeaways in my latest article. I hope it sparks some great ideas for your team! [Link] #CONVEXSummit #MySpeakerJourney #Volosoft #TechCommunity
-
-### 8. SEO Keyword Suggestions
-
-1. `CONVEX Summit 2026`
-2. `Natural language to SQL pipeline`
-3. `Alper Ebiçoğlu speaker`
-4. `Model Context Protocol MCP`
-5. `Abstract Syntax Tree SQL validation`
-6. `Dynamic database schema discovery`
-7. `Green software demand shifting`
-8. `Plain Concepts Madrid`
-9. `B2B software architecture AI`
-10. `ABP Framework database reporting`
-
-### 9. Social Media Hashtag Suggestions
-
-- `#CONVEX2026`
-- `#SoftwareArchitecture`
-- `#DotNet`
-- `#ConversationalData`
-- `#EnterpriseAI`

docs/en/Community-Articles/2026-06-25-Convex-Summit-2026-Recap/Post.md

@@ -10,98 +10,64 @@
 
 **SEO slug:** `convex-2026-ai-enterprise-reality-speaker-experience`
 
-[Convex Summit 2026 Developer](https://www.convexsummit.com/) conference held in the capital city of Spain, Madrid for the first time. [Plain Concepts](https://www.plainconcepts.com/) company organized this event. 2 days with 3 parallel sessions. 2 of the sessions were Spanish and one of them was English. It was organized on 17, 18 June 2026.
+Hi, I'm here for another conference reviews. This time I attended [Convex Summit 2026 Developer](https://www.convexsummit.com/) conference, held in the capital city of Spain, Madrid. It was my first talk in Madrid. The organizer [Plain Concepts](https://www.plainconcepts.com/) arranged a 2 days conf with 3 parallel sessions. 2 of the sessions were in Spanish and one of them was English. The conf dates were 17, 18 June 2026. It was in a very big cinema. In Lithuania I also spoke in a cinema, I guess it's better to arrange a conf in a cinema because of asthenosphere, acoustic and state visibility for visitors.
 
-There is a small moment before every conference talk when the room becomes quiet, the slides are ready, and you suddenly remember why the topic matters.
-
-For me, that moment happened at CONVEX Summit 2026 in Madrid.
-
-I was there as a speaker, presenting my session **“Chat with Your Data: Turn any database into a conversational reporting engine.”** But I was also there as a listener, a software architect, and someone trying to understand where enterprise AI is really going after the first wave of demos and experiments.
-
-CONVEX was interesting because it brought software development, architecture, and AI into the same conversation. These topics are often discussed separately, but in real companies they are tightly connected. AI features do not live in isolation. They live inside applications, databases, identity systems, permission models, workflows, dashboards, and business expectations.
-
-That was the real theme I felt throughout the event.
-
-![My CONVEX 2026 speaker badge before the sessions started](images/IMG_20927.jpg)
-
-*The speaker badge made the event feel real before the session even started.*
-
-![The CONVEX 2026 main stage in Madrid](images/IMG_20891.jpg)
-
-*The stage setup reflected the scale of the event: three communities, one shared conversation.*
+There is a small moment before every conference talk when the room becomes quiet, the slides are ready, and you suddenly remember why the topic matters. For me, that moment happened at Convex. My topic was “Turn any database into a conversational reporting engine.” I can admit it was a cool talk. But I was also there as a listener, a software architect, and someone trying to understand where enterprise AI is really going after the first wave of demos and experiments. CONVEX was interesting because it brought software development, architecture, and AI into the same conversation. These topics are often discussed separately, but in real companies they are tightly connected. AI features do not live in isolation. They live inside applications, databases, identity systems, permission models, workflows, dashboards, and business expectations.  That was the real theme I felt throughout the event.
 
 ## Speaking at CONVEX
 
+![My CONVEX 2026 speaker badge before the sessions started](images/\IMG_20927.jpg)
+
 My talk focused on a question that is becoming more important for enterprise software teams:
 
-**What if users could ask their business data questions in natural language—and receive safe, useful, validated answers?**
+> What if users could ask their business data questions in natural language and receive safe, useful, validated answers?
 
-I started my slide with Sobrino de Botin restaurant. It's the oldest restaurant in the world according to the Guinness Book of World Records. Sobrino de Botín **has been open since 1725**. The taste never changed, the restaurant **is keeping the same classic taste with the same cooking techniques**, but in the age of AI; we developers should **use new techniques for reporting**. We are not running a restaurant and this is how the development works. Adjust to the tech standards, new techniques and best practises.
+I started my slide with **Sobrino de Botin restaurant**. It's the oldest restaurant in the world according to the Guinness Book of World Records. Sobrino de Botín **has been open since 1725**. The taste never changed, the restaurant **is keeping the same classic taste with the same cooking techniques** but in the age of AI; we developers should **use new techniques for reporting**. We are not running a restaurant and this is how the development works. Adjust to the tech standards, new techniques and recent practices.
 
 ![image-20260624200005106](images/the-oldest-restaurant-world.png)
 
-And in one part of my talk, I need to ask questions like “Show me the top customers by revenue this quarter.” An AI model generates SQL. The system runs it. The user gets a result. And I learnt some Spanish sentences before the conf. Even for this I used AI. So I translated 10 diffferent English sentences to Spanish. Later I asked ChatGPT: "Can you score my pronunciation for these sentences.". ChatGPT gave the most ratings to 3 of my sentences' prouncations. And I talked those during my interview with my reporting AI tool. And that was fantastic eye-catching moment of my talk.
-
-
+And in one part of my talk, I need to ask questions like “Show me the top customers by revenue this quarter.” So the LLM answers in my language. And I learnt some Spanish sentences before the conf. Even for this I used AI. So I translated 10 different English sentences to Spanish. Later I asked ChatGPT: "*Can you score my pronunciation for these sentences.*". ChatGPT gave the most ratings to 3 of my sentences' prouncations. And I talked those during my interview with my reporting AI tool. And that was fantastic eye-catching moment of my talk.
 
-![Opening slide for “Chat with Your Data” at CONVEX 2026](images/IMG_20976.jpg)
-
-*My session was about building an AI-powered reporting tool that can understand a database and turn questions into controlled reporting workflows.*
-
-![Presenting my “Chat with Your Data” session at CONVEX 2026](images/IMG_20977.jpg)
+![Some photos from my talk 1](images/my-pictures-1.jpg)
 
 *My session focused on building conversational reporting experiences without giving up control, validation, or security.*
 
 I asked the attendees how many of you have written SQL, created reporting screen, and %80 of people wrote SQL and created reporting UIs. And same amount of people also use .NET. 
 
-One of my slides showed the reporting problem in a very familiar way. Traditional reporting tools are powerful, but they can be rigid. In many companies, a new report still means writing SQL, designing a page, testing the output, deploying a change, and waiting for a developer to become available. The result is that business users wait, developers become the bottleneck, and valuable data stays trapped behind technical complexity.
-
-The approach I shared has several layers.
+![Some photos from my talk 2](images/my-pictures-1.jpg)
 
-First, the system needs **schema awareness**. The AI should understand tables, columns, relationships, keys, data types, and enum values. Depending on the system, it may also need safe sample values or extra DDL information. Without that structured context, natural-language reporting becomes fragile.
+> **AI can make data more accessible, but architecture must define the boundaries.**
 
-![A slide from my session about teaching the AI the database schema](images/IMG_20974.jpg)
-
-*One of the most important parts of the talk was schema and system-prompt design: tables, columns, relationships, keys, data types, enum values, and controlled rules for query generation.*
-
-Second, it needs **intent interpretation**. Users do not always speak in database terms. They ask business questions. A good reporting engine should translate business language into technical meaning without forcing users to know the data model.
-
-Third, it needs **safe SQL generation**. Letting an LLM freely generate and execute SQL is risky. The model should operate under strict rules: read-only access, no destructive operations, tenant-aware filtering, query limits, validation before execution, and clear logging.
-
-Fourth, it needs **useful output**. A raw result table is not always enough. Sometimes the user needs an Excel file, a chart, a summary, or a clarification question.
-
-![A session slide about turning data access into a reporting engine](images/IMG_20980.jpg)
-
-Most business users do not want to “build a report.” They want an answer. But as developers and architects, our job is to make that experience safe, explainable, and reliable.
-
-AI can make data more accessible, but architecture must define the boundaries.
+---
 
-## What I Learned From the English Sessions
+## What I Learned From the Other Sessions
 
-There were 3 parallel session tracks. One of them was completely English sessions. One of the reasons I enjoyed CONVEX was that the English sessions did not treat AI as magic.
+There was a track with completely English sessions. One of the reasons I enjoyed Convex was that the English sessions did not treat AI as magic.
 
 The strongest message I heard across different sessions was this:
 
-**AI is becoming part of real systems, and real systems have constraints.**
+>  **AI is becoming part of real systems and real systems have constraints.**
 
 And I can see, software development is rapidly evolving with AI agentic tools. 
 
-> We cannot say development is dead, but hand-made development is dead. 
+> **We cannot say development is dead, but hand-made development is dead.** 
 
 From now on we'll use our time less on typing and more on thinking about features, user experiences and robust infrastructure. 
 
-![hand-made-coding](images/hand-made-coding.png)
-
+**In the age of AI, writing code by hand is the software equivalent of sending a fax to prove commitment.**
 
+![hand-made-coding](images/hand-made-coding.png)
 
-For a while, many AI discussions were focused on what AI could generate: code, tests, text, SQL, documentation, designs. At CONVEX, the more interesting question was: 
+For a while, many AI discussions were focused on what AI could generate: code, tests, text, SQL, documentation, designs. 
+At CONVEX, the more interesting question was: 
 
 **What happens after AI generates something?** 
 
 - Who validates it? 
 - Who owns the decision? 
+- Can we say I don't know to a question about how that works!?
+- If something blows up or happens a data leakage, who is accountable for it?
 - How does it fit into the architecture? 
-- How do we prevent data leakage? 
 - How do we make it useful for the business instead of impressive for five minutes?
 
 ### Architecture is also a people problem
@@ -112,56 +78,86 @@ One session that stayed with me used the idea of the **Prisoner’s Dilemma** to
 
 *The architecture sessions connected technical decisions with incentives, collaboration, and long-term system health.*
 
-Another slide suggested practical ways forward: learn the business, understand the competition, avoid “big bang” changes, work incrementally, create options, make trade-off decisions with the business, and **become a business value creator** rather than someone who only responds to requests.
+Another slide suggested practical ways forward: learn the business, understand the competition, avoid “big bang” changes, work incrementally, create options, make trade-off decisions with the business and **become a business value creator** rather than someone who only responds to requests.
 
 I liked that message. Architecture is much more valuable when it helps the business create options, not when it only explains why something is risky.
 
-### Power, knowledge, and decision-making
+---
+
+### Power, knowledge and decision-making
 
-Another interesting thread was about power in organizations. One talk referenced **Power-With**, an idea I found useful because it shifts the conversation away from control and toward collaboration. The slides connected power with access to knowledge, authority, charisma, and the way decisions move through an organization.
+Another interesting thread was about power in organizations. One talk referenced **Power-With**, I found it useful because it shifts the conversation away from control and toward collaboration. The slides connected power with access to knowledge, authority, charisma and the way decisions move through an organization.
 
 ![A session slide discussing Power-With and organizational dynamics](images/IMG_20914.jpg)
 
 *Some of the most interesting moments connected architecture with people, incentives, and organizational reality.*
 
-This may sound less technical than a database, a framework, or a deployment pipeline. But in practice, many technical decisions fail or succeed because of organizational dynamics. A clean architecture can still fail if teams are not aligned. A promising AI feature can still fail if no one trusts the output.
+There are 2 powers:
+
+1. **Power-Over**: Conquer other people's mind, control, force and order for a work output.
+2. **Power-With**: Co-operate with others, use everyone's power into a big power, move together for a work output.
+
+This may sound less technical than a database, a framework, or a deployment pipeline. But in practice, many technical decisions fail or succeed because of organizational dynamics. A clean architecture can still fail if teams are not aligned. 
+
+> A promising AI feature can still fail if no one trusts the output.
+
+### So Let's Think What's Charisma at Work
+
+- **What's charisma actually?** Let me tell you my opinion; charisma is experience, wisdom, grace, listening more and speaking less, way of looking to life, trustability (reliability), being a role model and an inspiration to others.
+- **Why charisma is important?** It makes your words to be listened by other people **naturally** (without any need of dictation).
 
 One slide quoted the idea that knowledge workers “think for a living.” Another referenced Peter Naur’s **Programming as Theory Building**, where program text and documentation are not always enough to carry the most important design ideas. That felt very relevant in the age of AI-generated code. If code becomes easier to produce, shared understanding becomes even more valuable.
 
-### AI agents need more than task execution
+As a summary;  AI can often understand the architecture from the code. What it usually can't know reliably is **why** 
+
+- the architecture ended up that way; the design decisions
+- trade-offs
+- historical context
+- assumptions that exist mostly in the team's shared understanding rather than in the code itself.
+
+---
+
+### What AI Agents can do and cannot do?!
+
+One of the most thought-provoking slides at the conf explored the current boundaries of AI agents. Instead of asking whether AI will replace humans, it asked a more interesting question: **What can AI agents actually do today, what can't they do and which limitations might disappear over time?**
+
+The first column listed the things AI agents already do well. They can execute tasks reliably without getting tired, maintain context across long-running work, report progress, detect problems, follow established processes, generate alternatives, document their work, and scale by running thousands of instances simultaneously. In short, AI excels at **execution**.
+
+The second column focused on capabilities that are fundamentally human today. AI can perform the role of a manager, mentor, or teammate, but it cannot truly *be* one. It cannot be held accountable for its decisions, earn trust through years of shared experience, genuinely care about an outcome, feel the weight of failure, or belong to a team. It can disagree or refuse a request, but not from genuine conviction or personal values. These qualities come from human relationships, responsibility, and lived experience, not from generating the next token.
+
+The third column added an important nuance. It wasn't titled "Impossible," but rather "Cannot, but maybe will one day." Some capabilities are already beginning to emerge. Multi-agent systems can delegate work to other agents. Better memory and continuity may allow AI to build trust over time. Multimodal models are becoming better at reading context, and reinforcement learning is an early form of learning from mistakes. The point wasn't that these problems are solved, but that some of today's limitations may become tomorrow's capabilities.
 
-The AI-related sessions also made a useful distinction between what AI agents can do today and what humans still bring to teams.
+The overall message was refreshingly balanced. AI agents should not be viewed as human employees, nor should they be underestimated. They are exceptional at executing work at scale, but organizations are built on more than execution. Trust, accountability, judgment, conviction, and belonging remain deeply human qualities, for now.
 
-One slide compared capabilities such as being assigned a task, executing work, holding context, reporting progress, flagging anomalies, following processes, generating options, reviewing work, documenting outputs, working asynchronously, and scaling across many instances.
 
-But the same slide also pointed to harder human qualities: accountability, trust earned over time, judgment under competing priorities, reading the room, feeling the weight of failure, belonging to a team, and genuinely caring about the outcome.
 
 ![A slide comparing what AI agents can and cannot do](images/IMG_20935.jpg)
 
 *The AI agent discussion was interesting because it did not only focus on automation; it also highlighted accountability, trust, judgment, and team dynamics.*
 
-That is a healthy way to talk about AI. Not “AI will replace everything,” and not “AI is useless.” The real question is where AI can help a team, and where humans still need to own the decision.
+That is a healthy way to talk about AI. Not “*AI will replace everything*,” and not “*AI is useless.*” 
 
-### Charisma at Work
-
-- **What's Charisma actually?** Let me tell my opinion; The charisma is the experience, the knowledge, wisdom, elegance, listening more and talking less, way of looking to life, performing good on the responsibilities, trustability, being inspiring.
-- **Why Charisma is important?** It makes your words to be listened by other people.
+> **The real question is where AI can help a team and where humans still need to own the decision.**
 
 ### Better decisions need better records
 
-I also followed sessions about architecture principles and decision records. One slide explained principles as priorities, beliefs, guardrails, and a way to connect requirements to architectural decisions. Another used a simple architectural decision example: **Data Store per Service**, where each service owns its data and other services access it through APIs or events instead of direct database queries.
+I also followed sessions about architecture principles and decision records. One slide explained principles as priorities, beliefs, guardrails and a way to connect requirements to architectural decisions. Another used a simple architectural decision example: **Data Store per Service**, where each service owns its data and other services access it through APIs or events instead of direct database queries.
 
 ![A slide about architecture principles and decisions](images/IMG_20952.jpg)
 
 *Architecture principles were presented as guardrails that connect requirements, trade-offs, and decisions.*
 
-This connected nicely with another slide about ADRs. A minimal ADR, based on Michael Nygard’s format, includes a name, status, context, decision, and consequences. A more comprehensive ADR can also include related requirements, assumptions, constraints, options, reasoning, and trade-offs.
+Let me first define what's ADR:.. An **ADR (Architecture Decision Record)** is a short document that explains **an important technical decision, why it was made, and what the consequences are**. See the [Microsoft Document about ADR](https://learn.microsoft.com/en-us/azure/well-architected/architect-role/architecture-decision-record).
+
+> **Code tells you *what* the system does. An ADR tells you *why* it was designed that way.**
+
+This connected nicely with another slide about ADRs. A minimal ADR, based on Michael Nygard’s format, includes a name, status, context, decision, and consequences. A more comprehensive ADR can also include related requirements, assumptions, constraints, options, reasoning and trade-offs.
 
 ![A slide explaining the minimal ADR structure](images/IMG_20960.jpg)
 
 *The ADR discussions were a reminder that good architecture is not only about making decisions, but also about preserving the reasoning behind them.*
 
-For .NET teams, these ideas are practical. AI can sit on top of strong foundations around backend services, identity, data access, cloud integration, and enterprise applications—but it should not bypass them.
+For .NET teams, these ideas are practical. AI can sit on top of strong foundations around backend services, identity, data access, cloud integration, and enterprise applications, but it should not bypass them.
 
 If anything, AI makes good engineering discipline more important.
 
@@ -169,45 +165,33 @@ If anything, AI makes good engineering discipline more important.
 
 The venue, Kinépolis Ciudad de la Imagen in Madrid, gave the conference a different feeling from a typical hotel-based event. The rooms, stage, and screens made the sessions feel cinematic. Outside the session rooms, the networking areas were active throughout the day. People were not only exchanging LinkedIn profiles; they were continuing the technical debates from the talks.
 
-![Participants networking at CONVEX Summit 2026](images/IMG_20888.jpg)
+![Participants networking at CONVEX Summit 2026](images/convex-ambiance.jpg)
 
 *The networking areas were busy between sessions, creating space for conversations beyond the formal agenda.*
 
-![Attendees moving between sessions at the venue](images/IMG_20890.jpg)
-
-*The event had a steady flow between talks, networking areas, and informal conversations.*
-
-For me, the most valuable conversations happened after the talk. Several people came with practical questions about AI-assisted reporting: How much schema information should be given to the model? Should generated SQL be shown to users? How do we prevent dangerous queries? Can this work with multi-tenant applications? How should we evaluate the quality of AI-generated reports? What should be logged for auditing?
-
-These questions matter because they show that teams are moving from curiosity to implementation. They are no longer asking only, “Can we do this?” They are asking, “How can we do this safely inside our product?”
-
-That is a much better question.
-
-![A personal moment from the event floor at CONVEX 2026](images/IMG_20970.jpg)
-
-*I'm preparing my setup and waiting the attendees from lunch :) One of those small personal moments that makes a conference feel memorable, not only useful.*
-
-A good conference gives you two things: ideas from the stage and better questions from the people you meet. CONVEX did both.
+I met people from my country as well from Bosch and Aselsan companies...For me, the most valuable conversations happened after the talk. I made new friends and learn what other people do.
 
 ## My Key Takeaways
 
 1. **AI features need data boundaries.** The more natural the interface becomes, the more important permissions, context, and allowed actions become.
-2. **Natural language is becoming a product interface.** Users increasingly want answers, not navigation paths.
-3. **Architecture is becoming more important, not less.** AI can accelerate delivery, but it cannot remove product constraints, security requirements, or organizational complexity.
-4. **Decisions need memory.** Principles, ADRs, trade-offs, and exceptions help teams preserve reasoning.
-5. **Conferences still matter.** A hallway discussion after a session can sometimes teach you more than a full article or video. It's a way of motivation, a way of socializing for developers. You see what other do, you discuss with them, you know your customers...
+2. **Natural language is becoming a product interface.** Users increasingly want answers, not get lost in the UI and not navigation paths.
+3. **Architecture is becoming more important, not less.** AI can accelerate delivery, but it cannot remove product constraints, security requirements, or organizational complexity. 
+
+   > When calculator first invented they didn't think problem solving finished with this invention, people use more time to solve problems rather then calculating....
+4. **Decisions need memory.** Principles, ADRs, trade-offs and exceptions help teams preserve reasoning.
+5. **Conferences still matter.** A hallway discussion after a session can sometimes teach you more than a full article or video. It's **a way of motivation**, a way of socializing for developers. You see what other do, you discuss with them, you know your customers and you know where you're at development.
 
 
 
 ## My Cultural Visits
 
-I visited Toledo and Madrid's most important tourist attractions and museums. Now I know way much better then I knew before about Spanish culture and lifestyle. But this is the most impressive moment for me. As you may know I'm Turkish. My grand grandfathers were Ottomans living in Anatolia. In 1571 they were in a war with Spain, Genoa, Malta and Italy. The battle was in the sea near Greece. It's called Sea Battle of Lepanto. In the below pictures, you can see the Ottoman's highest level sea commander (we call him Kaptan-ı Derya -the captain of seas-) personal items when he was died in this war. For those who wants to see it; it's Royal Palace of Madrid and the items are in Royal Armoury department.
+I visited Toledo and Madrid's most important tourist attractions and museums. Now I know way much better then I knew before about Spanish culture and lifestyle. But this is the most impressive moment for me. As you may know I'm Turkish. My grand grandfathers were Ottomans coming from Mongolia to Anatolia. In 1571 the Ottomans were in a war with Spain, Genoa, Malta and Italy. The battle was in the sea near Greece. It's called Sea Battle of Lepanto. In the below pictures, you can see the Ottoman's highest level sea commander's personal items. Wwe call him **Kaptan-ı Derya** -the captain of seas-. He died in this war. For those who wants to see it; it's in Royal Palace of Madrid and the items are in Royal Armoury department.
 
-If you're interested in this war, I'll little bit tell you about it.
+If you're interested in this war, you can also read this part:
 
 ### The Battle of Lepanto ⚔
 
-The Ottoman army was invading Cyprus. Angered by this, European countries asked Spain—one of the strongest kingdoms of the period—for help. To retake the island, a Holy Christian fleet was assembled under Spanish leadership. On 7 October 1571, the forces arrived at the Gulf of Patras in Greece for what would become known as the Battle of Lepanto. On one side was the Ottoman army, commanded by Ali Pasha, known as Kaptan-ı Derya (the Grand Admiral). On the opposing side were the major European powers: Spain, Venice, the Papacy, Genoa, the Knights of Malta, and Italy. It would become the last major naval battle fought with oared warships. The Ottomans lost the battle, and the sultan of that time Sokollu Mehmed Pasha said the following famous saying: “*You have cut off our beard, but we have cut off your arm. A beard grows back.*”  More than 200 Ottoman ships were lost. Tens of thousands of soldiers were killed or captured. In the below photograph, you will see war trophies taken from Ali Pasha. In Spain, this battle would be called “*La Defensa de la Cristiandad*” means “*The Defense of Christianity*” and would be used for propaganda for years. These trophies were used as symbols. The first time I saw the exhibit in a museum, I stood in front of it for 15–20 minutes, simply looking at it.
+The Ottoman army was invading Cyprus. Angered by this, European countries asked Spain, one of the strongest kingdoms of the period for help. To retake the island, a Holy Christian fleet was assembled under Spanish leadership. On 7 October 1571, the forces arrived at the Gulf of Patras in Greece for what would become known as the Battle of Lepanto. On one side was the Ottoman army, commanded by Ali Pasha (the Grand Admiral). On the opposing side were the major European powers: Spain, Venice, the Papacy, Genoa, the Knights of Malta, and Italy. It would become the last major naval battle fought with oared warships. The Ottomans lost the battle and Sokollu Mehmed Pasha (he's normally Serbian Turk and he's the most powerful manager after Sultan) said the following famous saying: “*You have cut off our beard, but we have cut off your arm. A beard grows back.*”  More than 200 Ottoman ships were lost. Tens of thousands of soldiers were killed or captured. In the below photograph, you will see war trophies taken from Ali Pasha. In Spain, this battle would be called “*La Defensa de la Cristiandad*” means “*The Defense of Christianity*” and would be used for propaganda for years. These trophies were used as symbols. The first time I saw the exhibit in a museum, I stood in front of it for 15–20 minutes, simply looking at it.
 
 ![royal-palace-0](images/royal-palace-0.jpg)
 
@@ -230,21 +214,19 @@ From the outside, it looks like nothing more than a “brutal spectacle,” but
 - **Bulls don’t react to the color red:** They are actually color-blind. What triggers them is the movement of the cape, not its color. Red is purely for visual aesthetics.
 - **The selected bull is a special, wild breed:** The animal has almost never seen a human before entering the arena. It sees the matador as an enemy and wants to destroy him.
 - **The greatest honor is to survive:** If a bull shows exceptional nobility and courage, the audience waves white handkerchiefs to ask for its pardon. That bull never enters the arena again and spends the rest of its life like a king on a farm.
-- **A dance with death:** Matadors do not see this as a sport, but as a way of confronting death. When making the most critical strike, the matador must also put his own life at risk—he cannot simply stab the bull from behind. He has to face the bull head-on, bravely. In that sense, there is a strange bond of respect between them. The bull is powerful, but the matador is intelligent. He cannot defeat it through strength, only through skill, timing, and agility.
+- **A dance with death:** Matadors do not see this as a sport, but as a way of confronting death. When making the most critical strike, the matador must also put his own life at risk, he cannot simply stab the bull from behind. He has to face the bull head-on, bravely. In that sense, there is a strange bond of respect between them. The bull is powerful, but the matador is intelligent. He cannot defeat it through strength, only through skill, timing, and agility.
 - **It is a controversial subject**, but hearing it firsthand in its own context changed my perspective considerably. At a time when animal rights are more important than ever, this tradition still continues. And I have now learned that bullfighting has a much deeper philosophy behind it than I had realized.
 
 ![bull-fight](images/bull-fight.jpg)
 
 ## Closing
 
-I left CONVEX 2026 with new ideas, useful feedback, and a stronger belief that the next phase of enterprise AI will be less about impressive demos and more about trusted systems.
+I left Convex 2026 and Spain with new ideas, useful feedback, and a stronger belief that the next phase of enterprise AI will be less about impressive demos and more about trusted systems.
 
 For me, the most interesting AI features are not the ones that look magical. They are the ones that quietly solve a real problem, respect the architecture around them, and help users make better decisions.
 
-Thank you to the CONVEX organizers, the speakers, and everyone who joined my session or continued the conversation afterward.
+Thank you to the Convex organizers, the speakers, and everyone who joined my session or continued the conversation afterward.
 
 Madrid was a great place to talk about AI, .NET, architecture, and the future of enterprise software. I hope to see many of you again at the next event.
 
-
-
 ---