AI & Systems
How I design and build real AI-powered systems, including autonomous driving and other reliability-critical environments: what works, what breaks, and how I fix it.
Hussam Ahmed
Building large-scale systems by day, exploring the universe by night.
I share practical insights on AI system design, technical execution, and astrophotography workflows, grounded in real builds, real constraints, and real outcomes.
Guide
What You'll Find Here
A quick map of the ideas, systems, and workflows this site is built to document.
Execution at Scale
What actually happens inside large-scale engineering programs and how decisions get made.
Astrophotography
Experiments, failures, and end-to-end workflows, applying system-level thinking to deep-sky imaging.
Writing
Latest Posts
Breakdowns, experiments, and observations from building and running real systems.
Jun 21, 2026
Loop Engineering for AI Coding Agents
Loop engineering is the practice of giving an AI coding agent a trigger, goal, execution mode, verifier, review gate, state log, and stop rule.
Read articleJun 18, 2026
Make Your Coding Agent Work Like Fable 5: A Step-by-Step Guide
Most of Fable 5's quality came from the order it worked in, not its weights: it read before editing, checked after editing, and changed course when a tool result broke the plan. That order shows up in session logs, so you can measure it and move it onto the model you already use — with seven copy-paste prompts, a CLAUDE.md playbook, and a test hook.
Read articleJun 7, 2026
Claude Code Dynamic Workflows: A Practical Guide to the New Orchestration Feature
A practical guide to Claude Code dynamic workflows: what the new feature does, when to use it, how to trigger it, and how to design workflows that split, verify, loop, and synthesize real engineering work.
Read articleBuilds
What I'm Building
Current systems and tools I'm shaping, with the objective, status, and friction points that matter right now.
AI & Systems
Map Knowledge Graph for Autonomous Driving
Autonomous Driving Prototyping
- Objective
- Interactive prototype showing how a map knowledge graph layer can reason about lanes, actors, crosswalks, and constraints.
- Current status
- Cinematic V1 simulator with scenario switching, graph traversal, timeline playback, and counterfactual reasoning.
- Key challenges
- Keeping the demo technically credible while making scene reasoning legible in seconds.
AI & Systems
AI Knowledge Hub
Agentic Coding · CLI & IDE agents
- Objective
- Interactive, fact-checked constellation of guides for mastering coding agents — live terminal sessions, config builders, and decision frameworks.
- Current status
- 9 constellations live: 89 interactive, fact-checked guides across Claude Code, Codex, Claude + Codex, Gemini CLI, IDE Agents, Foundations, The Toolkit, Microsoft Copilot, and Kiro.
- Key challenges
- Keeping pace with fast-moving agent tooling while never shipping an unverified claim.
Astrophotography · Systems Engineering
AstroBrain
- Objective
- A personal observatory planning model that turns fourteen nights, a real rig, a real horizon, and honest uncertainty into an explainable session.
- Current status
- Working PWA · explainable planning · anonymous first
- Key challenges
- Ephemeris, local obstruction geometry, lunar interference, weather provenance, optical fit, and immutable session history in one traceable system.
AI & Systems
Agentic Trading System
- Objective
- Autonomous system for market analysis and execution.
- Current status
- Prototyping live evaluation loops, guardrails, and decision logic.
- Key challenges
- Data pipeline reliability, model behavior under drift, and risk control under real conditions.
Execution at Scale
Program Management Systems
- Objective
- Automation for planning, execution tracking, and decision support.
- Current status
- Shaping workflows that turn program signals into clearer operating decisions.
- Key challenges
- Normalizing ambiguous updates, exposing tradeoffs, and keeping the loop useful at scale.
Execution at Scale
Productivity Tools
- Objective
- Internal tools to improve execution speed and visibility.
- Current status
- Iterating on small utilities that remove friction from day-to-day execution.
- Key challenges
- Finding the narrowest useful workflow, avoiding overhead, and keeping adoption effortless.
Astrophotography
Astrophotography Control Stack
- Objective
- System to control and automate imaging hardware across devices.
- Current status
- Connecting acquisition workflows across mount, camera, guiding, and capture software.
- Key challenges
- Integration reliability, automation across mixed hardware, and recovery from device edge cases.
About
Building real systems that have to work.
I work on systems outside controlled environments — incomplete data, moving requirements, and constant production pressure.
That means dealing with trade-offs and conflicting priorities while still delivering something reliable.
This site breaks down how I design, execute, and solve problems in that context.
Hussam Ahmed
Systems & AI · Technical program leadership
- Focus
- AI & Systems · Execution at Scale · Astrophotography
- Currently building
- Map Knowledge Graph for Autonomous Driving
- Based on
- Real builds, real constraints, real outcomes
Connect
Follow My Work
I share insights on AI systems, execution strategies, and applied engineering.
If you're building complex systems or working with AI, follow along.