Mentorship on how to build a Document Intelligence Platform

From Zero to Deploy

Build your own Document Intelligence Platform in 4 weeks. Document parsing with YOLO11, multi-model OCR and production deploy. One session per week in March/2026, with ongoing mentor support throughout the program.

Gabriel Chaves & Leandro Barbosa

AI Engineers | Document AI, RAG & Distributed Systems

Why this mentorship is different

Most RAG tutorials skip the most critical part: document ingestion. Without a robust parser, your RAG doesn't work. This mentorship solves that.

Tutorials that use generic loaders and ignore the real complexity of documents

Courses that skip OCR, layout detection and table extraction

Fragile pipelines that break with scanned PDFs or varied formats

Lack of guidance for async deploy with real monitoring

This mentorship covers everything: layout detection with YOLO11, multi-model OCR with smart fallback, async pipeline and production deploy on Render.

4 weeks with weekly sessions + ongoing mentor support between sessions.

What you will learn

FastAPI & Async Fundamentals

Async project setup with FastAPI, Uvicorn, middleware patterns and development environment configuration

YOLO11 Document Detection

Layout detection with YOLO11 (18x faster than Docling), 11 DocLayNet classes, global cache and parallel page processing

OCR Multi-Model Fallback

Fallback system with 5 models (FREE→Budget→Premium), 100% success rate, optimized prompts and 50% cost reduction

Async Pipeline & Task Orchestration

Hybrid Executor (ThreadPool + ProcessPool), Redis task management with WATCH-MULTI-EXEC, SSE streaming and webhooks

Storage & Caching

Cloudflare R2 integration, Redis caching (91.7% hit rate), batch processing and 11+ document format support

Production Deploy

Docker, deploy on Render, structured logging, Sentry for error tracking and production health checks

Mentorship Format

4 weeks in March/2026. One session per week + ongoing mentor support between sessions.

4 live sessions

One per week in March/2026

2h per session

Time for theory and practice

Ongoing support

Ask questions between sessions

Small cohorts

Guaranteed individual attention

Guided hands-on

You build alongside the mentors

Source code included

Complete repository for reference

4-Week Track

Week 1: Fundamentals

FastAPI async, document formats, OCR basics and project setup

Week 2: Detection & OCR

YOLO11 strategy, multi-model OCR fallback, region extraction

Week 3: Pipeline & Integration

Hybrid Executor, Redis tasks, R2 storage, SSE streaming, webhooks

Week 4: Production Deploy

Docker, deploy on Render, Sentry, structured logging and health checks

Your Mentors

Gabriel Chaves

AI Engineer | LLMs, RAG & Agentic Systems

Software and AI Engineer with strong experience in intelligent systems and agentic architectures. Designed conversational agents, NLP pipelines and model routing systems. Experience with LLMs, RAG, embeddings, vector databases, LangGraph, CrewAI and ADK.

Focus

LLMs, RAG, Multi-Agent Systems, Applied NLP

LinkedIn GitHub

Leandro Barbosa

CTO & Lead Engineer | AI & Distributed Systems

Lead engineer and CTO with extensive experience in large-scale distributed systems and AI infrastructure. Architected the largest Video CDN in South America (Globo). Built an ecosystem of 70+ AI projects including distributed inference systems and document intelligence platforms.

Focus

Distributed Systems, Production Document AI, Cloud & DevOps

LinkedIn GitHub Site

Who this is for

Ideal participants

Python developers who want to build data pipelines for AI

Backend engineers who need to process documents at scale

Professionals implementing RAG who need robust ingestion

Tech leads evaluating document intelligence solutions

Any dev with Python background wanting to evolve into AI Engineering

Not suitable for

• Complete beginners to programming
• Those looking for no-code or low-code solutions
• People without basic Python and async familiarity
• Anyone expecting results without committing to all 4 weeks

Prerequisites: Python knowledge and familiarity with REST APIs. Basic understanding of Docker and Git. No prior experience with OCR or YOLO required.

Frequently asked questions

No! Week 1 provides complete leveling. YOLO11 is used as a detection tool — you don't need to train models. We focus on integration and the pipeline.

No. This mentorship focuses on the document ingestion layer — the step before RAG. You'll understand how to prepare data for any RAG system.

Yes! On Week 4 you deploy your Document Intelligence Platform to production on Render with Docker, Sentry and monitoring.

Python with FastAPI for the backend. Docker for containerization and Render for deploy. Redis for caching and task management.

Yes, all sessions will be recorded and made available for participants to review.

R$ 150.00. Mentorship in small cohorts with direct support, complete source code and system deployed to production.

Ready to build your Document Intelligence Platform?

4 weeks in March/2026 with ongoing support. Limited seats.

Applications will be reviewed on a rolling basis. You'll hear from us within 48 hours.