Loading profile…
Daniil Makeev
Looking for a jobSenior Applied AI Engineer
I build reliable production LLM agents, automated evaluations, and resilient AI pipelines.
ArmeniaSenior Applied AI Engineer at Hao.vc
About
At hao.vc, I rebuilt an AI video-generation pipeline that cut production time from about 60 to 10 minutes and cost from about $2 to $0.80 per video. The system shipped 100+ fully AI-generated clips and helped project channels collect about 800K views.
I build LLM systems that can run unattended in production, including retries, fallback assets, media validation, typed recovery, automated evaluations, and repair of malformed model outputs. I also published Browd, a browser AI agent with human-in-the-loop permissions, six failure-recovery classes, an LLM-as-judge evaluation harness, and 218 automated tests.
I've also delivered a FastAPI and PostgreSQL interview-scheduling service for about 700 students, set up its CI/CD and deployment workflow, and built Python data automation that saved an estimated 8–12 hours of manual work each week.
Experience
Senior Applied AI Engineer
Wrote the generation engine of a production AI video platform; with the team moved it onto Prefect: turnaround 6x (~60 to ~10 min), cost -60% ($2.00 to $0.80). Redesigned the agent architecture of an open-source browser agent (LangGraph.js, Chrome MV3) with LLM-as-a-judge evals. Built contact-memory RAG over pgvector (DB calls 3,701 to 22) and a Telegram Mini App (FastAPI) for ~700 students.
PrefectLang ChainLangGraphPydanticAIAWSSupabaseDeep Agents
Python / Data Automation Engineer
Designed and maintained Python and SQL data-automation pipelines for catalog import/export, data validation, normalization, and internal operational workflows, saving non-engineering teams an estimated ~8–12 manual work hours per week. Built the backend and ETL data foundation for later applied AI and retrieval integration.
PythonSQLAutomationETLData Validation
Education
Bauman State Technical University