Hey! I'm
I build and ship production software — fast. A senior full-stack engineer with an AI-accelerated workflow that turns weeks of delivery into days.
I founded and led a software company — a team of ~15 engineers at peak. With a modern stack and an AI-accelerated workflow, I now deliver that same scope with 2–3 people. I've run it both ways.
~15 engineers→same scope→2–3 people
Available for contract and fixed-scope work — remote, across EU and US hours.
// what I sell
An agreed scope at an agreed price, working and deployed in days rather than months.
LLMs and automation inside a product people actually use, not a demo that impresses once.
Making the systems a business already runs on talk to each other, and stay in sync.
Senior engineering judgment on the days you need it, without a permanent hire.
// track record
~15 → 2–3engineers for the same delivery scope
18 yrsshipping production software, since 2008
100+projects delivered
1000+features shipped
2008–2018
Grew it from the first paying client to a ~15-engineer team, owning architecture, hiring and delivery across the whole portfolio — and the client relationships that funded it. Ran the business side and the codebase at the same time.
2018–now
Day-to-day delivery on software that real operations depend on: offline-first field apps, AI document pipelines, computer-vision workflows and the integrations that keep separate systems in sync. Production, not prototypes.
2020–now
Fixed-scope builds, AI integration sprints and fractional tech-lead work, taken on directly. Small scopes, agreed up front, shipped.
// proof
A few things I build a lot, and how they actually work.
capture→local queue→reconcile→sync
Work is captured straight to the device and dropped onto a durable queue, so a dead connection never loses anything. When the network returns, queued events replay in order and conflicts resolve against server state — the person using it never has to think about "offline."
TypeScript · IndexedDB · service workers · REST / JWT
ingest→extract→classify→validate→report
Inbound files move through isolated serverless steps: OCR pulls the text, an LLM classifies and structures it, results are checked against reference data, and a clean report comes out the other end. Each step is retryable, so one bad input never stalls the batch.
Python · serverless · Claude API · OCR
image→on-device model→vision-LLM fallback→structured result
A fast on-device model locates and segments objects and text in a photo in real time; anything it's unsure about escalates to a vision-LLM for a second opinion. Raw pixels go in, structured and checkable data comes out.
YOLO · on-device CV · vision LLM · TypeScript
scan sources→filter→LLM-score→dedup→digest
A scheduled job sweeps multiple sources, cuts the noise with a keyword pass, then has an LLM score each item for fit with a short reason. Everything seen is remembered so nothing repeats, and a single ranked digest lands in your inbox each morning.
serverless cron · Claude API · KV store · email
pull sources→narrow to active→aggregate→generate→schedule
A scheduled service pulls from several internal APIs, retries the ones that flake, narrows to the accounts that actually moved in the window, and writes the report the business runs its week on. The narrowing is the part that matters — a report covering everything is a report nobody opens, so it covers what changed and says plainly when nothing did.
Python · scheduled jobs · REST APIs · retry/backoff
reference prices→match items→price→snapshot→serve
Takes a reference price table and a set of items, works out what each one costs, and serves it behind an API other systems can query. The decision worth repeating is the snapshot: it is built deliberately free of personal data, so prices can be cached, shared and queried without dragging PII along behind them.
Python · FastAPI · SQL · pytest
source API→transform→dedup + retry→destination
Scheduled bridges that pull records from an internal API and push them to a third-party platform — format conversion, deduplication, retries, structured logging — so systems that were never meant to agree stay in sync without anyone babysitting them. Each one is small enough to read in a sitting, takes its credentials from a secrets manager with an environment fallback, keeps no state it does not need, and ships with a mock of the far end so the failure modes can be tested without touching anyone else's production.
serverless · AWS Lambda · Terraform · queues · REST / SOAP · webhooks
// the one you can actually open Most of my work sits behind an NDA. This one doesn't: flekit.com — a civic map of small neglected places, built and running in public. Source is open: github.com/lukasjandacom/flek — Cloudflare Workers, D1, R2 and a Vue 3 PWA, 288 tests, no dependencies.
// also: document & PDF generation · data-import pipelines · legacy API gateways · admin dashboards · environment sync tooling · accounting exports
// how I work
talk to users→requirements→sketch workflow→align→prototype→deploy→real use→tune
I talk to the people who'll actually use the thing, agree the workflow before anyone writes code, then get a functional prototype deployed and in real use as fast as possible. Everything after that is tuning against reality instead of guessing — which is why scope stays honest and nothing gets built for six months before someone notices it's wrong. Tasks and use cases get documented as I go.
// stack
// say hey
Got something that needs building?
→hey@lukasjanda.comUsually replies the same day. Happy to start with a 20-minute call.