Leonardo Barretti
AI / LLM Engineer — Production LLM Systems
I design and ship production-grade AI systems — multi-agent pipelines, RAG, and self-correcting workflows with real observability and cost control. 15+ years bridging business and technical execution in finance and trading.
Featured Projects
Cronograph
In ProductionProduction platform that extracts high-resolution market data from Binance and runs statistical window analysis to support weekly Bitcoin options strike selection. Sub-50ms aggregation over hundreds of thousands of OHLCV candles.
InvoiceReader
In ProductionStateful extraction pipeline in LangGraph that validates every LLM output against a strict Pydantic schema. Multi-LLM fallback with cost-aware model routing, field-level retries, and real-time SSE progress.
AuditChain
In Production5-agent forensic analysis system for SEC filings, orchestrated with LangGraph. Achieved 100% recall on known-fraud evaluation set using quantitative models (Beneish M-Score, Altman Z-Score) and RAG-based qualitative analysis.
bitPredict
In ProductionMulti-timeframe Bitcoin forecasting powered by Kronos, a 102M-parameter foundation model (HuggingFace). 30 stochastic simulations per candle with calibrated uncertainty bands and a portfolio backtest engine reporting Sharpe, drawdown, and win rate.
1.About Me
I'm an AI/LLM Engineer who designs and ships production-grade systems built around large language models — multi-agent pipelines, RAG, and self-correcting workflows with real observability and cost control. I pair deep domain understanding with hands-on engineering to take systems from prototype to production, where reliability, cost, and correctness matter.
Backed by 15+ years bridging business and technical execution in finance and trading, I bring a unique blend of stakeholder communication, system design, and pragmatic engineering. I care about code quality, testing (pytest), monitoring, and making systems that hold up under real load.
Open to opportunities in AI Engineering — available to relocation.
2.Tech Stack
Other Projects
RAG Systems
Retrieval-augmented generation pipelines with pgvector, citations, and streaming responses.
Doc Pipelines
Document-processing pipelines with LLM extraction, classification, and structured output.
Workflow Automation
AI-driven workflow automation for international clients — requirements to deployed products.