Blog
102+ articles from real AI-first delivery and two decades of building — deep dives first, full archive below.
Deep Dives
Beyond Code: A New Artefact and a Second Pipeline in the AI-DLC
Sep 2026As the AI development lifecycle matures, code stops being the only first-class artefact — and one delivery pipeline stops being enough. The latest chapter in the AI-DLC series.
#AI-DLC#AI-Native SDLC#Engineering Systems
AI Coding Made Us Faster. Then the Real SDLC Problems Started
Jul 2026Reflections from a first real AI-native project built with Claude Code and Codex: when coding becomes fast, the bottleneck moves to requirements, testing, UAT, and ownership.
#Claude Code#AI-Native SDLC#Human in the Loop
From AI Coding to AI SDLC: What Four Iterations Taught Us
May 2026Four iterations into a real AI-first project: why AI agents are like very capable (but imperfect) developers, and why traditional engineering discipline matters more, not less.
#AI-First SDLC#Claude Code#Engineering Discipline
Fewer Developers, More Builders (Part 2): The First Week of Going AI-First
Feb 2026Week one of a deliberately AI-first project: same AI, different people, different results — why AI-first development is a capability shift, not a tooling shift.
#AI-First#Cursor#Claude Code
Fewer Developers, More Builders: Why We're Starting an AI-First Project
Feb 2026Why we committed a real internal product to a deliberately AI-first SDLC — and why the future needs fewer traditional developers but many more builders.
#AI-First#Builder Mindset#SDLC
Ctrl+Alt+Delete My Career? A Veteran Developer's Take on How LLMs Are Changing Software Engineering
Jun 2025The 2023 reflection that started the journey: a veteran developer's early, cautiously optimistic take on how large language models would reshape software engineering.
#LLM#Career#Software Engineering
Archive · 96 articles
2026
- Geek Out Time: Agent Harness Exploration (Part 1) -What Happens When Success Is Impossible?Jul 2026
- Geek Out Time: Can AI Do Research by Itself? I Tried AutoResearch on Singapore Hawker FoodApr 2026
- Geek Out Time: Gradient Boosted Trees(GBT) Using a Hallucination Risk PlaygroundFeb 2026
- Geek Out Time: Tree-of-Thought for LLM ReasoningFeb 2026
- Geek Out Time: Running FastMCP with DeepSeek on Google ColabJan 2026
- Geek Out Time: Reject-by-Design -Determinism in Enterprise AIJan 2026
- Geek Out Time: From Ontology to AI Agents -Determinism Matters in High-Compliance IndustriesJan 2026
2025
- Geek Out Time: Why RAG Alone Fails Enterprise Governance (And How Ontology Fixes It)Dec 2025
- Geek Out - Why Enterprise AI Quietly Depends on OntologyDec 2025
- Geek Out Time: Can a Model Detect When It’s About to Hallucinate?Dec 2025
- Geek Out Time: The Economics of LLMs -How Token Pricing Quietly Shapes the ArchitectureDec 2025
- Geek Out Time: Testing TPU vs GPU on Google Colab After Meta’s Reported Shift Toward TPUsNov 2025
- Geek Out Time: Where LLMs “Keep” Their Facts? (VSCode with Colab)Nov 2025
- ICF ACC Exam Prep: Reflections and Lessons LearnedNov 2025
- Geek Out Time: DeepSeek-Powered Query Expansion and Self-Querying in Advanced RAGOct 2025
- Geek Out Time: When Geometry Fights Back — Why Your Embeddings in RAG Can’t Think in ‘AND’Oct 2025
- Geek Out Time: Reducing Hallucination with DoLa in Google Colab (Microsoft Phi-2)Oct 2025
- Geek Out Time: The “Rich” Get Smarter in RAG-Continuing from “Embeddings”Sep 2025
- Geek Out Time: Back to When I First Asked “What Is a Container?” -CGroups and Namespaces on ColabSep 2025
- Geek Out Time: From Prompts to Programs -How Open-Source DSPy + DeepSeek Train Better Autogen…Sep 2025
- 𝗚𝗲𝗲𝗸 𝗢𝘂𝘁 𝗧𝗶𝗺𝗲: Embeddings -Continuing from “Chunk Happens” (RAG)Sep 2025
- Geek Out Time: LoRA Fine-Tuning new Gemma-3–270M on a Free Colab GPUAug 2025
- Geek Out Time: Building a Kid-Safe AI Agent with Multi-Layer Protection (TF-IDF & Content…Aug 2025
- 𝗚𝗲𝗲𝗸 𝗢𝘂𝘁 𝗧𝗶𝗺𝗲: Training AI Agents for Tool Use with Reinforcement Learning in Google…Aug 2025
- Geek Out Time: Human-in-the-Loop Safety for High-Stakes AI Agents (HumanLayer + GPT-4 on Google…Jul 2025
- 𝗚𝗲𝗲𝗸 𝗢𝘂𝘁 𝗧𝗶𝗺𝗲: Chunk Happens — Testing different Chunking Strategies for RAGJul 2025
- Geek Out Time: Automating UI Testing with GPT-4 + Playwright in Google ColabJul 2025
- Geek Out Time: Routing Meets Retrieval-Context-Aware Multi-Agent Orchestration with AG2 + FAISSJul 2025
- Geek Out Time: Supercharging Multi-Agent AI: Hybrid Routing with AG2(formerly Autogen)Jun 2025
- Geek Out Time: Demystifying vLLM’s KV Cache, Latency & Context Isolation for Faster LLMsJun 2025
- Geek Out Time: LLM Inference Exposed-How It Works And How to Make It Fast on Google ColabJun 2025
- Geek Out Time: LLM Hallucination Experiments — Reflection, Verification, and Ground Truth in…Jun 2025
- Geek Out Time: Playing with Prefix Tuning on Google ColabJun 2025
- Geek Out Time: Simulating LLM Short and Long Memory with FAISS, LangChain, and Google ColabJun 2025
- Geek Out Time: Understanding Inference-Time Reasoning in LLMs with Google Colab and NVIDIA NIMMay 2025
- Geek Out Time: PPO vs GRPO on Google Colab: Two Ways to Align Language ModelsMay 2025
- Geek Out Time: Part 2- Super Cheap Fine-Tuning DeepSeek-7B with QLoRA, RunPod, and Cursor for your…Apr 2025
- Geek Out Time: Building Your Cheap Custom Chatbot using Gradio with Fine-Tuned Models on Hugging…Apr 2025
- Fostering a Culture of Continuous Learning: Guid SessionsApr 2025
- Geek Out Time: Building an Agent2Agent Startup Coach with the newly released Google ADK(Agent…Apr 2025
- Geek Out Time: Experiment with AI-powered Cursor- Reflections from the EdgeApr 2025
- Geek-Out Time: RAG vs. CAG in Google ColabMar 2025
- Geek Out Time: Upgrade Previously-Built AG2 Financial Copilot (Powered by DeepSeek) into an…Mar 2025
- Geek Out Time: Trying newly released OpenAI’s Responses API with Web Search Tool in Google ColabMar 2025
- Geek Out Time: Building a Multi-Agent Financial Advisor Copilot with AG2 (formerly AutoGen)…Mar 2025
- Geek Out Time: Simulating Distributed Training on TPU & GPU in Google ColabFeb 2025
- Geek Out Time: “Vibe Coding” on Google Colab with OpenAI & DeepSeekFeb 2025
- Geek Out Time: Mixture of Experts(MoE) vs. CNN: A Google Colab ExperimentFeb 2025
- Geek Out Time: Knowledge Distillation in TensorFlow- Smaller, Smarter Models in Google ColabJan 2025
- Geek Out Time: Build Your Own Autonomous AI Agent Backed by the Top Open-Source LLM DeepSeek v3 and…Jan 2025
- Geek Out Time: AI Model Routing — Dynamically Choose Models Based on Question ComplexityJan 2025
2024
- Geek Out Time: AI in the Browser- Run WebLLM for Powerful, Local LLM ExperiencesDec 2024
- Geek Out Time: Exploring Opensource AnythingLLM — The All-in-One, Easy AI Platform for Local RAG…Dec 2024
- Exploring LoRA on Google Colab: the Challenges of Base Model UpgradesDec 2024
- Geek Out Time: Building an Interactive Career Coach with Google Colab and Synthesizing Agents Using…Nov 2024
- Geek Out Time: Simulating High-Bandwidth and Cache-Like Memory on Google Colab’s T4 GPUNov 2024
- Geek Out Time: Exploration of Model Pruning for Efficient DeploymentNov 2024
- Geek Out Time: Experimenting with FP32 vs. FP16 Quantization on Google Colab’s Free T4 GPUOct 2024
- Team Coaching SessionOct 2024
- Geek Out Time: Play with the newly released OpenAI’s SwarmOct 2024
- GeekOut time: Exploring Complex SQL Queries with Natural LanguageSep 2024
- Geek Out Time: Building an Interactive Career Coach with AutoGen GroupChatSep 2024
- Geek Out Time: Creating a Graph-Based FSM Group Chat with Autogen (not Autogen Studio)locally with…Aug 2024
- Geek Out Time: RouteLLM with Nvidia NIM, A Cost-Effective Solution for Integrating Multiple Models?Jul 2024
- Geek out time: Creating an Agent with Customized Skill in AutoGen BuilderJul 2024
- Geek out time: try LLM and Embeddings on Nvidia NIM with Node.jsJul 2024
- Geek Out Time: Creating a Local AI Agent on My Mac Using Autogen Builder with the Local LLM…Jul 2024
- Build a Continuous Learning Environment through Tech Brown Bag SessionsJun 2024
- Geek Out Time: Playing with a Machine Learning Pipeline with PyTorch and MLflowMay 2024
- Geek Out Time: Simple Local Testing of Llama 3 on Its Release, Gemma, and Mistral MistralApr 2024
- Geek Out Time: Play with LangChain 3 — Simulate Full RAG Locally with word2vec and GemmaApr 2024
- Geek Out Time: Build a Facade API for OpenAI API and Local LLM APIMar 2024
- Geek Out Time: Play with LangChain 2- locally with GemmaMar 2024
- Geek Out Time: Play with LangChainMar 2024
2023
- Geek Out Time: Run web browser app with local SLMsNov 2023
- QKV and Multi-head Attention in LLMNov 2023
- Geek Out Time: Play with AutoGenOct 2023
- Here at Constellar, we know how to hustle hard and have a blast doing it!Jul 2023
- Our first Cybersecurity and Data breach tabletop exerciseJul 2023
- First-time use codes generated by ChatGPTJul 2023
- Tune the OpenAI model with NodeJSJul 2023
- Run Alpaca 7B on my MacbookApr 2023
- A Starry Starry NightMar 2023
- How can we use AI in the MICE industry?Feb 2023