Corporate Systems

Enterprise software that survives audit, scales with the business, and integrates with what’s already there.

Business outcomes

  • Replace manual audit and compliance workflows with automated, regulator-friendly systems

  • Modernize legacy stacks (Python 2 → 3, monolith → microservices) without breaking production

  • Connect ERP, finance, and operational systems through event-driven sync — products, prices, data flow in real time

  • Add AI agents (chat, RAG, automation) to existing platforms safely — with evaluation pipelines that catch regressions before customers do

  • Move workloads between cloud and on-premise / bare-metal where it makes business sense (incl. air-gapped, sovereign)

  • Cut manual ops with agent-driven workflows; humans set direction, agents handle implementation and review

Track record

  • Google Distributed Cloud Hosted — DevOps for air-gapped, on-premise Kubernetes (sovereign / regulated workloads). 08/2025 – present

  • DesignWeltDeko — Production AI agents (RAG, image classification, content generation) with LLM/RAG evaluation in production; custom ERP ↔ Shopify event-driven sync. 12/2024 – 08/2025

  • SAP — Automated audit framework features; continuation of distributed self-audit platform handling ~12 TB/day across 30+ business units. 2019 – 2024

  • Mercedes-Benz — Battery test analytics; microservice optimisation; Azure cloud migration. 2023 – 2024

  • Skoobe — Python 2→3 and Java→Python migration; GraphQL gateway consolidating legacy APIs; AWS CI/CD. 2018 – 2021

How I work

Modern, agent-driven development is the default workflow. Humans define intent; agents handle code, tests, review, and merge. Evaluation pipelines are the gatekeeper, not manual human review. The Concept → Plan → Spec → Code → Feature discipline keeps architecture coherent across multi-agent workflows. See: Agent-Driven Development.

Stack

  • Backend: Python (FastAPI + Pydantic, FastStream, asyncio), TypeScript / SvelteKit

  • Data: PostgreSQL, MongoDB, ScyllaDB, ClickHouse, Redis; Kafka, Redis Streams; Spark / pySpark, Airflow

  • Infrastructure: Kubernetes (cloud or bare-metal, incl. air-gapped), Docker, Terraform; full observability (Prometheus, Grafana, Loki, Tempo)

  • AI / LLM: OpenAI, Claude, LangChain, LangGraph, PydanticAI, FAISS, pgvector, MCP servers, eval pipelines

  • Compliance & security: OAuth / SSO, audit trails, schema validation, PII redaction, data mesh patterns

Full CV and project history: Profile Vladislav Vorobev.