💡 Tips: Ali Suleyman TOPUZ

Ali Suleyman TOPUZ

Nov 8

Memory Is an Update Problem: Three Authors, Three Answers, One Disagreement

#aisecurity #agents #llmengineering #architecture

12 min read

Ali Suleyman TOPUZ

Nov 8

I Compiled Five .NET Agent Tutorials Against the Real SDKs: What Worked and What Was Invented

#csharp #agents #dotnet #modelcontextprotocol

20 min read

Ali Suleyman TOPUZ

Nov 8

Agent Memory That Learns, Forgets and Stays Permissioned: A Week of Evidence

#dotnet #llmengineering #dataprivacy #agents

36 min read

Ali Suleyman TOPUZ

Nov 8

Route First, Retrieve Second, But Never Trust the Router’s Confidence

#rags #knowledgegraph #dotnet #llm

22 min read

Ali Suleyman TOPUZ

Nov 8

What Breaks When Your Knowledge Graph Goes Stale: A Small .NET Drift Experiment

#llm #dotnet #agents #knowledgegraph

45 min read

Ali Suleyman TOPUZ

Nov 8

Why “Spec-Driven Development” Needs a Deletion Test

#llmevaluation #specdrivendevelopmen #aicoding #agents

13 min read

Ali Suleyman TOPUZ

Nov 8

When Agents Route Around the Block: A Sandbox and Egress Checklist for .NET Agents

#agents #aisecurity #dotnet #agenticai

15 min read

Ali Suleyman TOPUZ

Nov 8

Cutting Agent Token Costs in .NET: Tool Search, Output Gating and What the Numbers Really Say

#modelcontextprotocol #costoptimization #llm #agents

25 min read

Ali Suleyman TOPUZ

Nov 8

The Steering Wheel Problem: run.steer(), turn/steer and Who May Redirect a Running Agent

#llm #security #promptinjectionattac #agents

17 min read

Ali Suleyman TOPUZ

Nov 8

Five Memory Types, One Store: I Merged and Split Agent Memory and Measured the Damage

#agents #llmengineering #dotnet #architecture

30 min read

Ali Suleyman TOPUZ

Nov 8

Skills That Maintain Themselves: SkillOpt, AutoHarness, and Why I Do Not Trust Either Yet

#agents #claudecode #dotnet #llmops

18 min read

Ali Suleyman TOPUZ

Nov 8

Eval-Driven Development for a .NET Agent: Replaying Traces Instead of Re-Calling the Model

#agents #testing #dotnet #opentelemetry

23 min read

Ali Suleyman TOPUZ

Nov 8

The Jev Gold Rush Is a Contract Story, Not a Model Story

#softwareengineering #llm #agents #artificialintelligen

26 min read

Ali Suleyman TOPUZ

Nov 8

I Built the Harness Tkachuk Describes: Approvals Enforced Outside the Model, in .NET

#dotnet #security #llm #agents

23 min read

Ali Suleyman TOPUZ

Nov 8

More Agents, Smaller Gains: What Agensh, MAST and the Google Study Say About Your Orchestrator

#agents #multiagentsystems #architecture #llm

20 min read

Ali Suleyman TOPUZ

Nov 8

What a 6.4 Percent Error Rate in a Benchmark’s Answer Key Taught Me About Trusting Agent Memory…

#marchinelearning #retrievalaugmentedge #llmengineering #agents

17 min read

Ali Suleyman TOPUZ

Nov 8

The Hybrid Beats Both: Routing Only the Uncertain Decisions to a Bigger Model, in .NET

#dotnet #ollama #aiengineering #llm

16 min read

Ali Suleyman TOPUZ

Nov 8

Where the Tokens Actually Go: Cache Rent, Effort Levels and Routing in a Long Claude Code Session

#agents #dotnet #llm #promptcaching

17 min read

Ali Suleyman TOPUZ

Nov 8

MCP Just Got Skills Built In, Here’s What Changes for .NET Developers

#agents #mcps #architecture #dotnet

18 min read

Ali Suleyman TOPUZ

Nov 8

Durable Agents in .NET: Kill the Process and See What Survives

#agents #distributedsystems #dotnet #csharp

20 min read

Ali Suleyman TOPUZ

Nov 8

I Ran a Harness Eval Loop on My Own .NET Agent. Here’s What Changed and What Didn’t

#dotnet #llm #softwareengineering #ai

16 min read

Ali Suleyman TOPUZ

Nov 8

The DeepSeek Harness Teardown: What a Running Medium Series Taught Me About Plugin Lifecycles

#plugins #deepseek #architecture #agents

14 min read

Ali Suleyman TOPUZ

Nov 8

Agent Memory After the Hype: What Jev-Mem, KnowledgeX and BlogWriter Taught Me About My Own Grep…

#sqlite #ai #llm #dotnet

28 min read

Ali Suleyman TOPUZ

Nov 8

I Tracked Every Cost-Cutting Trick AI Teams Used This Week. Here’s What Actually Works

#dotnet #agents #claudecode #llmcostoptimization

12 min read

Ali Suleyman TOPUZ

Nov 8

What a Week of AI Safety Incidents Actually Taught Me About Building Agents

#agents #dotnet #aisafety #agenticai

15 min read

Ali Suleyman TOPUZ

Nov 8

I Read the New MCP Apps Spec So You Don’t Have To: Here’s What Changes for .NET Developers

#csharp #security #agents #dotnet

14 min read

Ali Suleyman TOPUZ

Nov 8

Harness Engineering Is the New Bottleneck, Not the Model

#claudecode #agenticai #llm #harnessengineering

14 min read

Ali Suleyman TOPUZ

Nov 8

Tail Rent: What Actually Happens to Your Agent’s Memory After a Week of Continuous Runtime

#agents #llmmemory #aicodingagent #dotnet

17 min read

Ali Suleyman TOPUZ

Nov 8

Your Agent’s Memory Shouldn’t Be a Vector Database. It Should Be a Markdown File.

#dotnet #claudecode #agentmemory #contextengineering

14 min read

Ali Suleyman TOPUZ

Nov 8

A Week Inside Jev: What a Non-Generative Decision Model Is Actually Good For

#ai #llm #softwareengineering #claudecode

19 min read

Ali Suleyman TOPUZ

Oct 4

Your Knowledge Graph Is Measuring Your Pipeline, Not Your Subject: A Replication Attempt

#agenticrag #agents #python #llmengineering

12 min read

Ali Suleyman TOPUZ

Oct 4

Three Signals Say MCP Is Not Always the Right Tool, and What to Reach For Instead

#llm #agents #contextengineering #architecture

16 min read

Ali Suleyman TOPUZ

Oct 4

Multi-Agent Systems Don’t Have a Consistency Problem. They Have a Database Problem.

#ai #agents #dotnet #distributedsystems

17 min read

Ali Suleyman TOPUZ

Oct 4

Harness, Framework, or SDK: I Audited My Own Agent Stack Against a Real Source-Code Study

#harnessengineering #llm #dotnet #architecture

14 min read

Ali Suleyman TOPUZ

Oct 4

My Eval Suite Was Lying to Me: Auditing the Judge, Not Just the Agent

#dotnet #ollama #agents #llmevaluation

13 min read

Ali Suleyman TOPUZ

Oct 4

Who Authorized the Agent? Building Real Identity and Delegation for AI Agents in .NET

#identity #dotnet #agents #security

12 min read

Ali Suleyman TOPUZ

Oct 4

The Real Cost Curve of Running Agents in Production, From Four Different Companies

#promptcaching #modelrouting #llmcostoptimization #agents

17 min read

Ali Suleyman TOPUZ

Oct 4

.NET + AI in Production: What Actually Works (A Field Survey)

#csharp #ai #nuget #architecture

14 min read

Ali Suleyman TOPUZ

Oct 4

State, Memory, and the Two Camps Fighting Over What Your Agent Should Remember

#csharp #dotnet #agenticai #agenticmemory

16 min read

Ali Suleyman TOPUZ

Oct 4

Three Microsoft Agent Framework Posts, One Working App: RAG, a Custom Agent, and a Workflow Graph…

#dotnet #microsoftagentframew #csharp #rags

12 min read

Ali Suleyman TOPUZ

Oct 4

Your Agent’s Memory Is Probably Wrong: A Practical Guide to Agentic Memory Design

#aisecurity #llmengineering #agenticai #dotnet

16 min read

Ali Suleyman TOPUZ

Oct 4

DeepSeek Harness vs Codex vs Claude Code: What I Learned Comparing Three Agent Harnesses

#agentharness #agenticai #deepseek #openai

13 min read

Ali Suleyman TOPUZ

Oct 4

Seven Ways to Wire an Agent Harness, and What Actually Changes Between Them

#orchestration #claudecode #agenticai #harnessengineering

18 min read

Ali Suleyman TOPUZ

Oct 4

Prompts Are Consumables, Evals Are Capital: Rethinking What Actually Survives a Model Upgrade

#llmengineering #softwaretesting #dotnet #claudecode

15 min read

Ali Suleyman TOPUZ

Oct 4

Your AI Agent’s Reasoning Isn’t as Private as You Think

#apisecurity #llmsecurity #aisecurity

13 min read

Ali Suleyman TOPUZ

Oct 4

Building a One-Person Software Factory: Token-Efficient Agentic Workflows That Actually Hold Up

#developertools #claudecode #agenticai #llmcostoptimization

13 min read

Ali Suleyman TOPUZ

Oct 4

When (Not) to Use MCP in .NET: A Practical Guide

#mcpserver #agenticai #csharp #dotnet

14 min read

Ali Suleyman TOPUZ

Sep 17

Five Rules for a Claude Code Agent That Runs Itself

#autonomoussystem #claudecode #softwareengineering #agenticai

12 min read

Ali Suleyman TOPUZ

Sep 17

The Week Three Real Security Incidents Happened to AI Agents, and What Each One Actually Teaches

#socialengineering #cybersecurity #aisecurity #promptengineering

14 min read

Ali Suleyman TOPUZ

Sep 17

The Real Economics of Running Claude Code Agents in Production

#agenticai #costoptimization #llm #finops

13 min read

Ali Suleyman TOPUZ

Sep 17

What a 17.5x Cost Gap Between Coding Harnesses Actually Teaches You

#aicodingagent #promptcaching #dotnet #llmcostoptimization

15 min read

Ali Suleyman TOPUZ

Sep 13

Building a Production-Grade Eval Pipeline for Your Agent, Not Just a Demo

#llmevaluation #microsoftagentframew #dotnet #agents

12 min read

Ali Suleyman TOPUZ

Sep 13

Microsoft.Extensions.AI’s New Failover Feature Has a Streaming Blind Spot

#ai #dotnet #llmops #streaming

1

12 min read

Ali Suleyman TOPUZ

Sep 13

Inside a Real Multi-Agent Claude Code Setup: Two Leads, 9 Projects, 40 Prompts a Day

#architecture #multiagentsystems #agenticai #developertools

13 min read

Ali Suleyman TOPUZ

Sep 13

Your .NET Agent Is in Production, Your Engineering Discipline Isn’t.

#microsoftagentframew #llmevaluation #dotnet #agenticai

13 min read

Ali Suleyman TOPUZ

Sep 13

I Got Claude Code to Talk to GPT, Grok, and a Local Ollama Model. Here’s What Actually Held Up

#ollama #agenticai #llmrouting #claude

8 min read

Ali Suleyman TOPUZ

Sep 13

Graph as Architecture, Not Graph as Data: Why I Stopped Reaching for GraphRAG First

#graphrag #agenticai #langgraph #rags

13 min read

Ali Suleyman TOPUZ

Sep 13

Claude Code Is an Operating Layer for Engineering Work: I Went Looking for Proof Instead of Just…

#claude #agenticai #developertools #mcpserver

13 min read

Ali Suleyman TOPUZ

Sep 13

The World Your Agent Thinks In: What Two Ontology Experiments Taught Me About Graph Design

#llmengineering #knowledgegraph #dotnet #agents

13 min read

Ali Suleyman TOPUZ

Sep 13

Agent Skills vs MCP: I Stopped Reading Hot Takes and Measured the Actual Context Cost

#architecture #llm #agenticai #mcpserver

12 min read

Four Year Club

Three Year Club

Two Year Club

Writing Debut

One Year Club