Skip to content

About

I came to AI engineering through software engineering, and it shows.

I’m Hussain Ahmad, an AI Software Engineer based in Faisalabad, Pakistan. I build agent systems, retrieval-augmented applications and the full-stack software that has to support them.

My background is in building conventional systems — APIs, access control, queues, realtime delivery, test suites, deployment pipelines. That turned out to be the useful preparation. Most of what makes an AI feature trustworthy is not model work: it is validation at the boundary, work moved off the request path, provenance carried with an answer, and a way to tell whether last week’s change made anything better.

So I treat a model as one component inside a system that has to keep its promises. I’m interested in where that component should be trusted, where it should be constrained, and where it should hand back to a person.

I work in the open. Everything on this site links to source, and where a result has not been measured I say so rather than filling the gap with a number.

Role
AI Software Engineer
Based in
Faisalabad, Pakistan · PKT (UTC+5)
Focus
AI agents, RAG, LLM applications, full-stack systems

What I look for

The work I want to be doing

Being specific about this saves everyone time, including me.

  • Problems with a real cost of being wrong

    Systems where an incorrect answer has a consequence are more interesting than ones where it does not, because they force the design questions that matter: confidence, escalation, auditability.

  • Work that reaches production

    Prototypes are useful and I build them, but the engineering I care about starts at the point where other people begin to depend on the thing.

  • Teams that write things down

    Architecture decisions, trade-offs and the reasons behind them. It is the difference between a codebase that can be changed and one that can only be added to.

Currently exploring

What I'm working through right now

Open questions I'm spending time on, rather than a list of things I already know.

  • Agentic AI

    Multi-step agents that plan, call tools and know when to hand back to a person.

  • AI evaluation

    Making prompt and retrieval changes measurable against a fixed question set instead of judged by feel.

  • Production RAG

    Ingestion freshness, retrieval quality and citation integrity at the point where a corpus stops being small.

  • Long-term agent memory

    What an agent should keep between sessions, and what it should be made to forget.

  • Multi-agent architectures

    When splitting work across specialised agents beats one capable agent, and when it only adds coordination cost.

  • AI infrastructure

    Queues, workers, streaming and observability — the layer that decides whether an AI feature survives contact with users.

Contact

Have an AI product or workflow worth building?

Send me the problem — not the spec. If it’s a fit I’ll tell you how I’d approach it; if it isn’t, I’ll say so.

Faisalabad, Pakistan · PKT (UTC+5) · Working remotely