Stanislav Rybonka

Solution architecture on every project — system design, data modeling, integrations and scaling — while staying hands-on for core and high-risk components. Technical discovery and pre-sales, engineering standards, hiring and mentoring a senior-only remote team.

• Dev13 Systems — architected and built an AI-native business OS end to end: 8 modules and 25+ Claude-powered specialist agents in production on a human-approval workflow, with MCP servers exposing platform tools to the agents
• Built the shared retrieval and evaluation layer behind every AI project — hybrid search over pgvector and Qdrant with reranking, Langfuse and LangSmith tracing, LLM-as-judge regression tests on golden datasets, per-tenant token and cost tracking
• Leon & Vera — AI lead-generation SaaS for European real-estate brokers over the WhatsApp Business API; live in 6 markets with subscription billing, GDPR and EU AI Act compliance

Technology strategy across a multi-domain platform — messaging, learning, marketplace and payments — plus the core backend services and the infrastructure they run on. Product direction, backlog and release planning with the founders.

• Designed the AI agent layer — assistant workflows, LLM tool calling across every platform module and task automation, driven by a single-button assistant
• Built real-time messaging across text, voice and video
• Delivered wallet and payments — token balances, Mastercard-powered card issuing, crypto-to-fiat conversion and settlement flows

Features end to end on an enterprise whistleblowing platform used by Fortune 100 companies and government agencies — design, implementation, tests and deployment — plus architecture decisions across platform services.

• Ran enterprise onboarding implementations for large customers
• Mentored engineers, ran code reviews, owned team performance and development plans
• Reported status, risks and decisions to internal and client-side stakeholders

Application design, architecture and technical direction on a wealth-management platform for financial advisors. Risk management, feature estimates and release planning, continuing with the Steelkiwi team after Globaldev's 2021 acquisition.

• Helped scale the engineering team from 5 to 20+ — interviewing, onboarding and technical ramp-up
• Contributed to ISO 27001 certification — security controls in application architecture and data handling
• Refactored the legacy internal CRM — structural and architectural rework plus database optimization

A Python CRM platform for an Australian fintech — client management, authentication, granular permissions and secure data processing. Grew from developer to Tech Lead over four years on the account.

• Built data synchronization handling 500K+ records per day across third-party financial integrations
• Cut the bug rate by 30% and shortened delivery cycles by tightening the SDLC — review standards, automated testing and release process
• Extended the core engine with Stripe, Xplan, Google and AWS integrations; built payment processing and financial data flows

Backend features on a crypto-trading monolith with real-time peer-to-peer transactions — the first production system where I owned features rather than tickets.

• Backend development in Python and Django across the trading platform
• Unit testing with pytest, and working requirements into defined scope with the client