- Find me on Twitter 🐦️
- Read my occasional technical articles 📜
- For business inquiries: r83b7u4af@relay.firefox.com 📬️
I'm an entrepreneurially minded engineer that like to takes ideas from concept to completion. Building startups has helped me refine my fullstack dev skills and consulting for big fintech companies has helped me get a better understanding of ops side and best SRE practices, both in terms of systems building as well as human interactions with the system.
Recent work:
- Founder & CTO - Langolin - a language learning app
- Built a custom NLP engine to help with tasks such as alignment of phrases, quality check, etc
- Designed a very low cost yet robust infra without any significant outages
- Node/Express, MongoDB/Redis, Nginx, Python/Tensorflow/Torch
- SRE & Consultant - a German publicly listed fintech
- Part of on-call team
- Helped bring best practices to increase velocity and bring visiblity to metrics that matter
- Kubernetes, Prometheus/Grafana, Thanos/Cortex, Terraform, Azure/GCP
Intersection of AI 🤖️ and Games 🎮️ is what I'm particularly interested in. I used to play Poker and Go but nowadays I play Dota 2 occasionally.
Interestingly, all three mentioned games have been disrupted by breakthoughs in ML in recents years:
- Superhuman AI for multiplayer poker
- Self-taught AI is best yet at strategy game Go
- Dota 2 with Large Scale Deep Reinforcement Learning
I've tried replicating the works above in following repos:
- Node.js with Koa/Express for backend and React for frontend
- Python for general AI/ML as well as occasional server purposes
- Rust for high-performance code
- And others depending on use case