Writing

Field notes on AI engineering, multi-agent LLM pipelines, RAG, text-to-SQL, and shipping production software.

• Aug 5, 2026 · 3 min read · GodForge

  The execution gap: why GodForge won't let an LLM touch your files

  GitHub Copilot suggests code, then leaves you to copy, paste, and run the rest of your workflow yourself. GodForge — my VS Code extension — closes that gap with deterministic logic for anything risky, and hands the LLM exactly one job. It's now a peer-reviewed, published paper.

• Aug 4, 2026 · 3 min read · Kunaal's Portfolio — Ask AI

  Chrome ships an LLM now — no API, no server, no rate limit

  Gemini Nano runs inside Chrome itself as a browser capability, not a service you call. It isn't a hosted API and it isn't 'install Ollama and run it yourself' either — the browser downloads and manages the model, and any site can use it for free. Here's what building a real feature on it is actually like.

• Jul 24, 2026 · 3 min read · Repository File Count

  Exploring an enterprise repo without cloning it first

  Why 'just clone it and look' doesn't scale for large open-source or enterprise repos — and how Repository File Count answers what's actually inside one in seconds, straight from the GitHub API.

• Jul 21, 2026 · 2 min read · AI Interview SDK

  Testing an AI interview for bias before a real candidate sees it

  AI Interview SDK ships a Bias & Consistency Harness and an Interview Simulator — scripted personas, including an adversarial prompt-injection attempt, run against your real rubric in your own CI.

• Jul 15, 2026 · 2 min read · AI Interview SDK

  A follow-up engine that's actually dynamic

  Inside AI Interview SDK's core: a depth-limited, repeat-preventing, difficulty-scaled follow-up engine that reacts to how the candidate is doing — not a fixed question list.

• Jul 8, 2026 · 2 min read · AI Interview SDK

  Shipping an open-source AI interview SDK that stores nothing

  How AI Interview SDK drops an AI-scored interview into any app as one React component — with your keys, your backend, your database — so the maintainer never sees a candidate's answer.

• Apr 18, 2026 · 2 min read · Query Forge AI

  Schema-aware text-to-SQL: prompting LLMs to write correct queries

  Lessons from building Query Forge AI and SQLumina — why naive natural-language-to-SQL hallucinates, and how feeding the schema makes LLM-generated queries actually correct.

• Mar 4, 2026 · 2 min read · AI Flowchart Studio

  Inside a 4-stage multi-agent pipeline (FastAPI + Gemini)

  The architecture behind AI Flowchart Studio: a 4-stage multi-agent LLM pipeline — Orchestrator, Logic Parser, Generator, Validator — that turns plain English into rendered diagrams.

• Feb 12, 2026 · 2 min read · CommitAI

  Building CommitAI — a local-first AI git assistant with Ollama

  How I built CommitAI: an offline, zero-key git assistant that turns staged diffs into clean Conventional Commits using a local LLM via Ollama and Gemma.