Loading...
Loading...
Technical lead in agentic AI, still shipping code
I started as a backend engineer on services that 51M+ people hit every day, and I still write the core code. Today I lead the architecture of a production AI agent platform at Alien Intellect: agent memory, RAG with real citations, tool calling, guardrails, and the evals that decide when a change is good enough to ship. Along the way I grew the engineering team from 2 to 11 and built the habits that let a young product ship on a fixed rhythm. I start from the why and the what, then build the how, and I own the bar for done.
Alien Intellect
Took the company's AI agent from concept to private beta in small measured steps, owning both the product direction and the architecture. Led the migration of the production agent from five separate contexts into one unified architecture, shipped its persistent per-user memory and a RAG pipeline that answers from a user's own documents with real citations, and built the AI guardrails that keep the agent away from its own prompts, source code, and infrastructure. I own prompt and context engineering and the eval suite that gates every agent change. Alongside the product I grew the engineering team from 2 to 11 in a year, wrote the hiring rubric and two-stage interview flow, run onboarding and coaching for every joiner, and built the team's working culture from zero. When the summer 2026 war weeks hit Iran, I made attendance optional, replanned around outages, kept the daily as the one fixed point, and the July beta still landed on its date.
Cafe Bazaar
Owned architecture for products serving 51M+ users, including the payment platform. Designed a 10,000+ RPS system that cut average response time 65% while holding 99.9% uptime. Led the Python-to-Go migration across 15+ services with zero downtime, reducing memory usage 75% and improving response times 80%. Drove a microservices consolidation that cut service count 40%, inter-service calls 60%, and infrastructure costs 30%. Built the automated documentation system later adopted across the company, and pushed a complete-ownership culture that took deployments from monthly to daily and cut MTTR 70%. Also drove early LLM adoption: cut user-review response time from about 2 days to under 15 minutes and rolled out Claude Code across engineering.
Independent
Advise 2-3 early-stage startups part-time, alongside full-time roles, on architecture decisions, scalable systems, and the processes that keep working as a company grows.
Ferdowsi Cloud
Built Django REST APIs integrating with OpenStack and wrote Python monitoring services for cloud infrastructure. This gave me hands-on experience with distributed systems and early cloud adoption, and it still shapes how I think about infrastructure today.
Ranked first out of 200 students in Computer Engineering
Master's in Computer Engineering at Iran's top-ranked technical university
2-year membership in Iran's National Elite Foundation
Architected and scaled distributed systems handling 10,000+ requests per second
Architected the payment platform serving 51+ million users at Cafe Bazaar, with a 99.99% transaction success rate and zero critical security incidents
Peer-reviewed research in Blockchain, Healthcare IoT, and Fog Computing (Springer, IEEE, Scientia Iranica)
I believe in the "zero to 100" philosophy - complete ownership of the entire product lifecycle. Every problem starts with the why and the what before the how. As a Technical Product Manager who still codes, I write production Python (with some Golang), manage CI/CD pipelines, self-manage Kubernetes deployments, monitor with Prometheus and Grafana, lead cross-functional teams, and make strategic decisions based on KPIs, OKRs, and A/B tests. That ownership extends to architecture itself: at every company I've worked with, the migration strategy, scalability, and integration decisions have been mine.
Deep optimization and performance are fundamental to both AI systems and traditional infrastructure. From architecting a unified AI agent with persistent memory and RAG to migrating Python to Go for 10,000 RPS scalability, technical excellence drives business success. I own prompt engineering and define the evaluation criteria that decide whether an agent's output is good enough to ship. It is the same discipline I brought to system profiling and benchmarking on the infrastructure side.
With several years of teaching experience and community organizing (Ferdowsi LUG), I've learned that the best engineers are also great teachers. That extends to hiring: I grew Alien Intellect's engineering team from 2 to 11 in a year, designed its interview process and scoring rubric myself, personally interview and evaluate every candidate, and mentor new engineers through onboarding. Developing documentation systems, mentoring teams, and contributing to open-source creates a culture of learning that elevates everyone.
When I'm not architecting systems or leading teams, I play sports regularly to stay active and balanced. I spent years organizing the Ferdowsi Linux User Group and volunteering in university and Linux associations, giving back to the community that helped shape my career.
Robotics, where I won multiple awards, still shapes how I think about systems. Years of teaching, from high-school robotics to Quera's Go and Python bootcamps for Cafe Bazaar, built the communication and storytelling skills that product leadership depends on.