# Dmitry Ivanov: AI Implementation Engineer

> Portfolio of 28 AI systems and agents shipped to production or as working prototypes: document processing, classification, RAG, voice agents, outreach, local LLMs, developer tooling. Interactive map of the projects at https://neurocraft.space/.

Dmitry Ivanov builds and ships applied LLM systems end to end: from prompt and schema design, evals and retrieval to deployment and monitoring. Status labels: "Production" means real users or real data; "Working prototype" means it works end to end but is not a full rollout.

## Pages

- [Agent map, English](https://neurocraft.space/): interactive map of all projects, filters by audience, status and area
- [Agent map, Russian](https://neurocraft.space/ru/): the same page in Russian
- [Agent map, Serbian (Latin)](https://neurocraft.space/sr/): the same page in Serbian
- [Full project details (Markdown)](https://neurocraft.space/llms-full.txt): every project with the key facts, the pipeline and the stack

## Projects in production

- SMS risk classifier: Spam, phishing and compliance scoring for a US SMS provider. (8M+ messages · 98.9%; Classification)
- Tender Processing: Tender emails in, classified tenders and ERP-matched line items out. (300–500 emails a day · 92%; Documents, Classification, Enterprise)
- Living Status: A project board that an LLM keeps accurate by reading what the coding agents report. (every commit tied to a card; Agents & infra, Knowledge & RAG)
- Meeting bot: Joins a Google Meet call (Zoom supported), writes the transcript and the summary. (in use at 2 companies; Voice & audio, Enterprise)
- RVC Global: A job search platform I co-founded in 2023: collects and classifies jobs, matches them to CVs. (10K+ users · 2,000+ jobs/day; Products, Classification, Data & monitoring)
- Bank statement processing: Statements from 5 banks in, typed transactions and a spreadsheet out. (96.3% pages error-free; Documents, Vision & images)
- Document processing: Finds the document type, then reads out the fields that type needs. (400 documents a day · 91%; Documents, Vision & images, Classification)
- Cold email outreach: From finding the company to the reply inbox, with a personal email for every contact. (25K+ contacts; Outreach & GTM, Data & monitoring)
- AI bot for helpdesk: Every ticket classified, enriched with account data and answered with a draft reply. (every ticket classified; Enterprise, Classification, Knowledge & RAG)
- Local LLM inference: On-premise inference with local models for several client systems. (Qwen text + vision · 2 GPUs; Agents & infra, Vision & images)
- Corporate AI chat: A private company AI chat with an anonymizer and an LLM proxy in front of the models. (local network · limits + logs; Agents & infra, Enterprise)
- News and event monitor: Parses chosen sources, tags every item and sends short digests. (digests + tags; Data & monitoring, Classification)
- Clarifier: Enriches product item cards: classification, properties and market data. (classify · enrich · compare; Data & monitoring, Classification)
- Price and product search: Stock and prices from chosen suppliers, plus search for matching products. (no AI needed · AI optional; Data & monitoring)
- Procurement approval system: Every external payment of an industrial group goes through it. (~1,000 users · 4K+ requests/mo; Enterprise)
- Service work orders: Field service tracked in one ERP system across many sites. (35+ sites · 1,000+ orders/mo; Enterprise)
- Public MCP server: Vacancy search that any AI assistant can call. (read-only tools; Agents & infra, Data & monitoring)

## Working prototypes

- BasBot: My own agent platform: planner, sandboxed executor, memory, and a model choice per stage. (per-stage models · sandbox; Agents & infra, Knowledge & RAG)
- AI Company Brain: Meetings, tickets and CRM joined into one place that answers questions and lists action points. (Meet · HubSpot · Jira; Knowledge & RAG, Voice & audio, Enterprise)
- Docs assistant (RAG): A RAG system over marketing content and an API reference, with an API that writes the answers. (RAG · answer API; Knowledge & RAG, Documents)
- Lead research tool: Company research, scoring and personalized drafts for a marketing team. (1,000+ companies; Outreach & GTM, Classification, Data & monitoring)
- Scan vs original check: Compares a scanned document with its original and highlights every changed place. (up to 80 pages; Documents, Vision & images)
- Contract review: Compares an incoming contract with the client’s own template, clause by clause. (clause checks + redlines; Documents, Knowledge & RAG)
- Ad creatives optimizer (PropellerAds): Finds weak ad creatives in PropellerAds campaigns and proposes replacements. (control panel · 6-hour sweeps; Outreach & GTM, Vision & images)
- Call analytics: Call recordings turned into transcripts, speakers and checklist-based findings. (speakers + checklist; Voice & audio, Enterprise, Classification)
- Restaurant staffing + voice agent: A staffing product with a voice agent that onboards restaurant candidates and covers shifts by phone. (built in 2 weeks; Products, Voice & audio, Classification)
- Android non-custodial crypto wallet: A multi-chain wallet with swaps and bridges, first version in under 2 weeks with AI coding agents. (11 networks · swaps + bridges; Products)
- Language-learning apps: Four apps for language teachers: lesson generator, audio lessons and a handwriting check. (~50 teachers · 13 languages; Products, Voice & audio, Vision & images)

## AI stack

- Coding agents: Claude Code Max, Codex Pro, Devin, OpenCode, Droid (Factory), Cursor, Grok Heavy
- Models: Claude, OpenAI GPT, Gemini, DeepSeek, Kimi, GLM, Qwen (local), JeV (typesafe.ai), OpenRouter
- LLM frameworks: LangChain, Langfuse, MCP, Schema-guided reasoning (SGR), Pydantic structured output, own agent runtime, GEPA (prompt optimization)
- Retrieval: pgvector, BM25 + vector hybrid search, FAISS, re-ranking
- Voice and vision: Whisper, ElevenLabs, Twilio, PaddleOCR, Tesseract, Docling, vision LLMs, YOLO
- Local inference: vLLM, Ollama, quantized Qwen

## Contact

- Email: sntpeter@gmail.com
- Telegram: https://t.me/dima_rvc
- GitHub: https://github.com/maketeamake
- Site: https://neurocraft.space/