Sergei Vasilev
Engineering Team Lead · Solution Architect · AI Engineer
Belgrade, Serbia · open to relocation in EU
About
I solve hard engineering problems — at any level of the stack.
Give me a business problem and I'll turn it into architecture. Give me an architecture and I'll build it myself. Over 11 years, I've taken 10+ products from whiteboard to production — working side by side with founders and CTOs to shape both the technical vision and the business logic behind it.
Currently Python Practice Lead and Team Lead at Dualboot Partners, delivering for US clients across real estate, healthcare, aerospace, and fintech. My work spans Temporal-orchestrated AI pipelines with OCR and RAG, to cloud-native infrastructure handling billions of records. Previously built Dualboot's Balkan Tech Center from 4 to 18+ engineers across Serbia and Montenegro.
Open to senior engineering roles, consulting, and projects worldwide.
What I do
Practice & team lead for cross-functional teams of 4–18. I set technical direction, define standards, run hiring, and keep delivery consistent — without losing touch with the code.
I translate business problems into architecture. Microservices, async APIs, cloud-native infra on AWS/GCP — designed to scale from day one, not bolted on later.
End-to-end AI pipelines: OCR, RAG, vector search, prompt orchestration, embeddings, evals. I build systems that process real documents, answer real queries, and run in production.
I work hands-on at every layer — backend, frontend, mobile, infra. Brought 10+ products from whiteboard to production, side by side with founders and CTOs.
Selected Projects
CRE intelligence platform serving 400+ US real estate organizations. Architected backend infrastructure processing billions of records — complex ETL across dozens of data formats, multi-source financial aggregation, and hundreds of thousands of reports per month.
Built end-to-end from 0: backend APIs, customer-facing web app, mobile app, and full cloud infrastructure. 350,000+ active users. Stripe-powered payments and subscription management.
LLM pipeline for a precision manufacturer supplying NASA, commercial aviation, and defence. Automated ingestion and parsing of technical specs, compliance documents, and engineering artifacts — replacing weeks of manual processing.
Temporal-orchestrated AI pipelines for high-volume batch processing of bankruptcy filings, mortgage docs, risk reports, and financial Excel data. OCR, multi-class classification, and structured extraction at scale.
Analyses codebases and business artifacts to auto-generate epics, tasks, user stories, and use cases. Cuts requirements-to-ticket cycle from days to minutes.
Backend for a US hospital network HR system — nurse recruitment, credential verification, and shift management across multiple hospital facilities.
Real-time odds engine with multi-wallet and multi-token support, automated settlement, and live event streaming. Built 0→1 with the founding team.