# Daniil Makeev — AI Engineer

> Production AI engineer building agent systems, RAG pipelines, and content generation stacks.

**Canonical URL**: https://wyddy.tech/

## Summary

- AI Engineer at Hao, building video generation pipelines and retrieval systems.
- Focuses on multi-agent architectures, retrieval quality, pipeline speed, and generation cost.
- Stack: Python, LangGraph, deepagents, Supabase, Docker, OpenRouter.

## Selected Work

### Hao — Video Generation Pipeline

- Rebuilt a full video generation pipeline: topic to published mp4.
- Dialogue, TTS, images, FFmpeg compose, S3 upload — two-stage JSON API.
- Results: **6× faster** throughput, **−60%** generation cost.
- Stack: deepagents, LangGraph, ElevenLabs, FFmpeg, OpenRouter, Docker.

### NetWho — AI CRM Assistant

- Telegram-native AI assistant over a personal professional network.
- RAG with pgvector, Jina AI news enrichment, reranking loop.
- Live demo with multi-tenant contact RAG, Supabase RLS, pgvector, and reranking.
- Public demo: https://t.me/netwho_bot
- Stack: deepagents, LangGraph, pgvector, Supabase, aiogram, Docker.

### Digest Bot — Personalized AI Digest

- Self-hosted Telegram bot running daily.
- Two-agent architecture: stateless pipeline + stateful conversational agent with LangGraph checkpointing over Supabase.
- Stack: deepagents, LangGraph + Supabase checkpointer, OpenRouter, python-telegram-bot.

## Contact

- LinkedIn: https://www.linkedin.com/in/wyddy
- GitHub: https://github.com/wyddy7
- Email: wyddy7@gmail.com
- Telegram: https://t.me/wyddy7

## Agent Resources

- Agent skills index: https://wyddy.tech/.well-known/agent-skills/index.json