| name | project-map |
| description | Project-to-stack map for Daniil Makeev's featured AI work. |
Project map
Use this skill when you need to connect featured outcomes to the underlying engineering work.
Hao - Video generation pipeline
- Outcomes: 6x faster processing, 60% lower generation cost, batch mode
- Technologies: LLM dialogue, ElevenLabs TTS, FFmpeg, Backblaze S3, Docker
- Main themes: orchestration, cost control, pipeline reliability
NetWho - AI CRM assistant
- Outcomes: live Telegram assistant over personal and organizational networks
- Technologies: RAG, pgvector, aiogram, Pydantic v2, Supabase, Docker
- Main themes: retrieval, context assembly, tenant isolation
NetWho - Retrieval layer
- Outcomes: grounded answers from source records with cross-tenant isolation
- Technologies: embeddings, reranking, Supabase RLS, multi-tenant schema
- Main themes: answer quality, source grounding, safe multi-user use