┌─ wwakabobik@github:~/profile ─────────────────────────────────────────────┐
│ tty1 · login: Ilia Vereshchagin · shell: zsh · uptime: 18y+ │
└───────────────────────────────────────────────────────────────────────────┘
Python expert. Avionics → silicon → fintech → streaming → multi-agent AI.
Same person who signed DO-178B Class A and ships LangGraph agents.
$ cat ~/mission.txtI fix the $100k+/month problem of slow release cycles and engineering idle time.
Build AI QA processes and frameworks from scratch — multi-agent LangGraph harnesses, pytest suites, and the CI/CD that ties it together — so product teams ship daily with confidence.
Equally comfortable architecting QA for a 1200-engineer org and writing the Python that runs it.
$ cat ~/open_to.txtStaff / Principal SDET · Head of QA Engineering · Fractional QA Leadership · consulting
$ history | grep -E 'employer|client' | grep -v toxic
Liebherr-Aerospace safety-critical fly-by-wire · Airbus / Bombardier / Sukhoi
Auriga × Intel hundreds of thousands of tests/day · years before Appium existed
Luxoft test strategy & audits · Deutsche Bank delivery centre
Mediaplatform 300 → 1500+ tests · daily DevOps pipeline · pylint 5.3 → 9.98
Statusmoney pytest/Selenium/Appium from zero · ~80% coverage · 7 platforms
Now multi-agent AI QA · frameworks from scratch · RAG · LLM-as-a-judge
$ tree ~/skills -L 1 --dirsfirst
$ journalctl -u career --priority=crit -n 8 --no-pager● hardcore.timeline
├─ Certified Class A avionics (DO-178B / MISRA) for Airbus, Bombardier, Sukhoi
├─ Intel R&D: mass-scale toolchain + early Android HW tests — pre-Appium era
├─ Mediaplatform: 5× test suite growth into a daily release pipeline
├─ Statusmoney: automation from scratch across 7 platforms (~80% coverage)
├─ Raised legacy framework health pylint 5.3 → 9.98 without freezing delivery
├─ Zero-to-one RAG marketing platform (LangGraph · Postgres · S3 · email agents)
├─ Productised AI bots: Telegram / WhatsApp / Discord + TTS + WebRTC voice
└─ Multi-agent AI QA from scratch: processes, harnesses, LLM-as-a-judge eval
$ ./bin/snake --from contribution-graph --render svg
# contribution cells get eaten. daily cron keeps it honest.
$ fortune | cowsay -f tux _________________________________________________________
/ Novi Sad · Remote EU/US · EN fluent · RU native · SR B2 /
| Amateur rocketry · astrophotography · board games · cook|
\ Best engineers still build things on weekends. /
---------------------------------------------------------
\ ^__^
\ (oo)\_______
(__)\ )\/\
||----w |
|| ||
┌─ session end ─────────────────────────────────────────────────────────────┐
│ logout · reconnect: linkedin.com/in/wwakabobik · wwakabobik.github.io │
└───────────────────────────────────────────────────────────────────────────┘