Field guide
A forward deployed engineer embeds where the problem actually lives — owning the outcome, not just the deliverable, and shipping until it holds. As AI grows more capable, that’s the work that compounds. This is a field guide to the craft: real builds that show what it looks like, and playbooks on how to build the skill.
Why this exists
Becoming a forward deployed engineer isn’t an incremental skill upgrade — it’s a fundamental change in how a technologist works. The traditional IT model — take a spec, hand back a deliverable, own none of the outcome — trains the wrong instincts for it. For anyone building toward this work, that model isn’t just unhelpful; it’s detrimental.
And as AI absorbs more of the write-code-to-spec work, there may be no durable technology career for most people that isn’t some form of forward deployed engineering: sitting with the problem, owning the outcome, shipping the thing. This platform exists to make that path legible — to show what the work is, and how to grow into it.
Production tools, each shown in full — the problem, the constraint that made it hard, the one architecture decision that mattered, and what’s still unfinished. What the work looks like in practice.
A study tool for FDE interview prep — describe any project in plain English and it generates sample FDE-style answers across 12 parameters in four tiers (is it worth doing, can it be built, how to deliver, will it survive production), each with 2–3 different defensible angles rather than one "correct" one. Runs with no key on curated free models, or on your own OpenRouter key for stronger ones.
A teaching-grade accounts-payable agent that pays clean invoices touchlessly, routes every exception to a human, and makes each judgment (deterministic / LLM / human) visible on screen.
A multi-account Jyotish app that computes a real sidereal birth chart with Swiss Ephemeris, then has an LLM read that exact chart — specific placements, not generic horoscope filler — with a model-comparison harness and an LLM judge to keep it honest.
A teaching-grade telecom next-best-offer agent — it picks the right offer for each subscriber under hard eligibility rules, pauses for a human on sensitive retention offers, invents and prices new offers behind a finance gate, and shows the cost and accuracy of every model call. Runs with no key on recorded data, or live on your own key.
A demo practice-management platform for an Indian CA firm where AI does the grunt work — GST reconciliation, ledger scrutiny, notice drafting — and every output is a draft that waits for the partner's sign-off.
Field notes on the craft: how the role actually works, the architecture traps to avoid, and how to build a career doing it.
Field notes by email
New builds and playbooks on forward deployed engineering — occasionally, when there is something worth sending. No spam; unsubscribe in one click.
Behind the guide
Written and maintained by Aviral Mathur, a forward deployed engineer building AI systems with enterprises. Who’s behind it, and how to reach me →