Description

I build production-ready AI on top of two decades of backend engineering. That hybrid is what the market now calls an AI Engineer: LLM orchestration, RAG and autonomous agents, wired into backend systems built to hold up under load.

What I help with:

• AI features that ship and stay up. Multi-agent orchestration on OpenAI Agents SDK and LangGraph, RAG, MCP tooling, evaluation and guardrails. I build SearchMap.AI (30+ agents, 130+ MCP endpoints across 8 services). Around the agentic core, an n8n automation layer handles the operational outer loop: scheduled jobs, deploy pipelines, health monitoring and alerting, data ingestion, external integrations. It's self-hosted, so the whole thing stays owned and auditable.

• Backend and distributed systems. High-availability microservices and multi-region active-active DR. At Kaleyra, a four-continent WhatsApp platform, that meant 99.9% uptime on 1M+ messages a day, with MTTR down from 45 minutes to under 10.

• Platform and modernization. Microservices migrations and microfrontend/platform architecture for enterprise SaaS (BMC, Fortune 500 customers), plus API design and CI/CD.

A note on the "AI engineer" label: it's everywhere now, and most of it is a year or two old. What I bring is the 20+ years of production judgment behind it. More than a thousand code reviews. Real systems under real traffic. A feel for what breaks at scale. The result holds up in production, not just in a demo.

Stack: Python, Java, Spring Boot, Go, FastAPI, LangGraph, OpenAI Agents SDK, MCP, n8n, RAG with Qdrant, PostgreSQL, Kubernetes, Docker, AWS, TypeScript and React.

EU citizen, Partita IVA. B2B only, EU remote. English C1, Italian C1.

Industry field of expertise

Languages

English
Fluent
Italian
Fluent
Russian
Native or bilingual

Workplace preferences

Remote only

Primarily works remotely

SearchMap.AI
Senior AI Engineer
October 2025 - Today (10 months)
Milan, Italy
Multi-agent research platform: search, crawling, content classification, structured report generation. 30+ specialized agents across 3 services, Python + Java + TypeScript, 8 containerized services.

- Dual agentic orchestration (OpenAI Agents SDK + LangGraph) with feature-flag canary rollout and A/B metrics
- 130+ MCP tool endpoints across 16 modules, enabling agent-to-agent workflows
- Production resilience: circuit breakers for LLM tools, workflow checkpointing, loop guards
- Java MCP server (Spring Boot 4, Spring AI, LangGraph4j): 9-node extraction pipeline with model escalation
- RAG pipeline: Qdrant vector search, 5 research scenarios, 6-agent research chain
- LLM-as-Judge evaluation at zero extra cost
- Event-driven: Redis pub/sub, SSE streaming, Celery workers. 34-page frontend (Next.js 16, React 19)
- n8n automation layer (self-hosted): scheduled jobs, deploy pipelines, health checks and alerting, data ingestion, external integrations

Stack: Python, Java, TypeScript, FastAPI, Spring Boot, LangGraph, Celery, PostgreSQL, Qdrant, Redis, Docker, GitHub Actions
Java Spring boot Python LLM Typescript
Kaleyra (Tata Communications)
Senior Lead Developer / Staff Engineer
December 2023 - November 2025 (1 year and 11 months)
Milan, Metropolitan City of Milan, Italy
Staff/Lead IC on enterprise WhatsApp messaging platform (Meta premier partner). Reliability, multi-region delivery, developer enablement.

- Active-active DR across several dozen Go microservices (2 regions, four continents) — 99.9% uptime on 1M+ messages/day, 13 architecture docs, 20+ engineers coordinated
- Resilience framework (RabbitMQ/MySQL/MongoDB): auto-recovery, graceful shutdown via Uber FX — MTTR 45 min → <10 min
- Pricing/rate card evolution (150+ countries, multi-currency) — 80% reduction in pricing errors
- 1,000+ PR reviews (peak 103/month), 37 repos, mentoring 26 engineers
- Shared Go library: reusable RabbitMQ + Redis Enterprise adapters across all services
- 12 internal recognitions (7 awards + 5 badges)
- AI enablement: integrated Cursor AI into dev workflow, promoted spec-first AI-assisted development

Stack: Go, Python, TypeScript, Uber FX, RabbitMQ, MySQL, MongoDB, Redis Enterprise, AWS, Kubernetes
GO Kubernetes Microservices RabbitMQ Amazon Web Services
BMC Software
Senior Product Developer
November 2020 - October 2023 (2 years and 11 months)
Milan, Metropolitan City of Milan, Italy
Architect and lead developer for enterprise SaaS capacity planning product (Fortune 500 customers).

- Microfrontend shell platform unifying 6 apps (JSP, AngularJS, Angular 14) via 3 integration strategies: iframe orchestration, postMessage, Module Federation
- Migrated queuing-network engine from monolith to Java 17 microservices; 5-7 REST APIs (OpenAPI-first)
- Built full UI for Cloud Migration Simulator (public-facing, 5 cloud providers)
- Led monorepo migration for all frontend applications
- L3 production support (~15-20 incidents over 3 years)

Stack: Java 17, Spring Boot, Angular 14, Module Federation, PostgreSQL, MongoDB, Docker, Kubernetes
Java Spring boot Docker Microservices Architecture Kubernetes