Hi, I'm
Mikhail Shytsko
Founder of Seedfast. Solutions & AI Engineer at Slovenská sporiteľňa, based in Bratislava, Slovakia. BS in Computer Science, Technical University of Košice (2025).
sporiteľňaSeedfast
Focused on creating value, not just writing code.
Building AI-native systems and the backends behind them: MCP-based agents, LLM pipelines and high-load services, with a strong Spring/JVM background now applied in Kotlin and Python. At Slovenská sporiteľňa, Slovakia's largest retail bank, I build the pipelines and AI tooling behind services serving millions of customers.
An AI-native platform generating large-scale synthetic datasets for compliance-sensitive industries — finance, pharma, medtech. It reads database schemas and produces realistic, relationship-aware data in under 3 minutes. Won 1st place at DDAccelerator Finals as the only Slovak startup competing across 9 countries.
Where I've shipped.
- Designed and built an MCP-based AI assistant that summarizes and cross-references lending data spread across several internal banking systems — replacing the manual, multi-system lookup that slows credit review.
- Architected the AI's data access as a least-privilege, fully auditable MCP tool layer, with strict data minimization and a controlled disable path — safe inside a regulated bank without altering the core approval workflow.
- Built change-detection that surfaces and explains material changes across products, limits, collateral and client data — so reviewers catch shifts they'd otherwise miss across systems.
- Delivered a six-figure cost reduction by replacing a legacy workflow with an in-house Kafka pipeline — production-ready, end-to-end from a one-page spec.
- Cut response times 85% for two high-load services serving 2M+ customers by taking sole ownership of inherited services and driving targeted performance optimizations.
- Built an AI system that generates realistic, referentially-correct data for any database — sparing AI, QA and dev teams days of manual data-engineering per dataset, in a single run.
- Hardened the generation core with automated correctness checks — eliminating a class of silent data-integrity failures.
- Built custom evaluation frameworks that made the non-deterministic generation pipeline measurable and guarded against regressions — used them to more than double its output quality on held-out data.
- Added locale-aware generation in any language, so output reads natively for any locale, not just English.
- Shipped the developer-facing Go CLI and an MCP server — letting engineers and AI agents run Seedfast in production, CI/CD and agentic workflows.
- Built a secure, EU GDPR-compliant backend from scratch on Spring WebFlux + GCP, encrypting sensitive patient data in real time.
- Automated a CI/CD pipeline that cut deploys from 12 to 4 minutes.
- Automated equipment certification, cutting document-certification time ~50%.
- Built GraalPy trend analytics that accelerated product decisions ~25%.
- Shipped a secure Spring Boot REST API with OAuth2 / JWT authentication.
- Cut shift-approval time from 5 days to 2 through schedule optimization.
Selected work.
A mix of products, platforms and developer tooling — built to be owned, not just demoed.
Background & highlights.
Get In Touch
Open to collaborations, interesting problems, or just a friendly chat.