AI Systems Consulting
Your AI worked on a demo, fell apart in production?
I help tech leads and founders launch AI systems that actually work.
Sound Familiar?
AI system requires constant babysitting
Two developers just fixing prompts instead of shipping features. AI was supposed to save time, but it's eating team resources.
"Sometimes works, sometimes doesn't"
Worked great on demo, hallucinates in production. System is like a lottery — hit or miss.
Tokens eating the budget
Budgeted $500/month for tokens, burned through $5K. No idea how to optimize without losing quality.
Impossible to debug
AI broke, don't know where to look. Logs exist, zero value. Black box instead of a system.
"ChatGPT can do it = we can do it"
ChatGPT showed great results, but API in production outputs something completely different. Expectations didn't match reality.
Recognize any of these? I help solve exactly these problems — systematically, without magic tricks.
How I Help
Two work formats for different needs:
AI Systems Consulting
I design AI system architectures that work in production. From initial design to implementation strategy.
What you get:
- Architectural plan with clear component boundaries
- Testing and observability strategy
- Integration roadmap with existing systems
- Risk assessment and mitigation plan
Good fit if:
- Building AI product from scratch
- Converting demo to production
- Need independent architecture expertise
- Legacy system + AI = fear of breaking things
Technical Advisory
Fractional CTO / Technical Advisor for startups. Regular technical expertise without hiring full-time CTO.
What you get:
- Strategic technical guidance
- Help with critical technical decisions
- Architecture and code review
- Participation in hiring and team building
Good fit if:
- Startup without CTO/Technical Lead
- Need regular technical expertise
- Building AI product, need experienced advisor
Client Reviews
RAG in our chatbot worked just terribly. We tried to optimize it in every possible way - all to no avail. At first I was hostile towards Ivan - in the first session he asked a lot of questions, and what we didn't have time to discuss, he asked me to fill in as answers in a checklist. While I was doing this myself, I started to guess that the problem wasn't with RAG, but that it simply wasn't needed for our task. In further work with Ivan, we rebuilt the system into a simpler one. As a result, it started working more accurately (according to tests) and became cheaper to maintain. Most importantly - we began to understand how it works, where to look if something breaks, and how to expand the system.
Our team was building an agent to process customer requests from mailboxes. At first everything worked fine, but then we found ourselves in a state of constant support of the bot itself and re-checking of formed requests. We were about to abandon it completely, but we would have had to hire more customer relationship managers. We turned to Ivan for consultation. Ivan looked at our system, talked with the team, and a few days later brought, as he modestly said - "an approximate PoC of how it should work." This PoC as is has been working for us for the third month now with practically no complaints.
We needed to develop an AI module for our old backend that's about 6 years old. We tried ourselves, but it worked with too large delays on requests, and it wasn't clear at all how it would live on. Ivan immediately said that it couldn't be done this way, and started asking uncomfortable questions. Honestly, I thought he was just stalling for time. But when he showed the diagram of how the integration should be - everything fell into place. We rewrote the AI service in a month, and now the system works stably. Tokens cost us about $400/month, although we had budgeted three thousand for this load. I liked that Ivan around his initial diagram wrote out decent documentation.
Projects
Production Agentic AI System @ Monite
Multi-agent AI system for financial platform with multi-stage pipeline, structured logging, and Schema Guided Reasoning
Technologies: Python, FastAPI, Pydantic, OpenAI, PostgreSQL, PgVector, semantic-router, Kubernetes, SGR
AI Insurance Document Processing Pipeline
End-to-end automated insurance document processing system with NLP, reducing claim processing time from 3-5 days to <10 minutes
Technologies: Python, FastAPI, Pydantic, PostgreSQL, Docker, LLM, SGR
Enterprise AI Platform for Employee Onboarding
RAG-based intelligent onboarding platform with evidence-based learning practices, reducing average onboarding time from 3 to 1 month
Technologies: Python, FastAPI, langchain, LLM, ChromaDB, fastembed, PostgreSQL
For Whom
You're my ideal client if:
✓ Tech Lead or Engineering Manager
Building AI system with a team, need expertise
✓ AI Startup Founder
Demo works, production breaks
✓ CTO/Technical Director
Need independent architecture expertise
When to reach out:
- ✓ AI system is unstable in production
- ✓ Need to design AI system from scratch
- ✓ Tokens eating budget, need optimization
- ✓ No observability — impossible to debug
- ✓ Legacy + AI — afraid to break working system
- ✓ Need independent architecture review
- ✓ Team drowning in AI support instead of shipping features
When NOT to reach out:
- ✗ Looking for a "magic prompt" that solves everything
- ✗ Need quick hacks — I'm about proper architecture
- ✗ Not ready to invest in system quality
Get in Touch
Fill out a short form (2 minutes), describe your situation. I'll send work format options and available time for a discovery call.
Or reach out directly: zakutnii.ivan@gmail.com
How We'll Work
- Discovery Call (free, 30 minutes)
Discuss your situation, see if we're a good fit - Scope & Proposal
Define scope, format, cost - Work
Consulting or advisory — depending on the task
What's Next?
Submit a request through the form or email me, briefly describe your situation (2-3 sentences), and we'll schedule a discovery call.
— Ivan Zakutnii
AI Systems Engineer
staff engineer @ octa
P.S. If you're not sure my help is right for you — reach out anyway. I'll honestly tell you if I can't help or recommend someone else.