Context before the model
Discovery, business rules and trustworthy sources give agents useful direction.
In 60 seconds
Applied AI with context, criteria and results.
See how I turn language models, knowledge and business rules into reliable solutions that reach production.
About
Applied AI, with judgment
My edge is not training models: it is specifying, building and evaluating GenAI systems that work in production. I come from Requirements Engineering, so every agent is born with use cases, business rules, guardrails and evaluation metrics.
I evaluate LLMs on international platforms (Turing, Outlier AI) with SxS and fact-checking, and my MSc research (UFPB) evaluates the quality of user stories generated by different LLMs, with blind analyst review.
How I turn AI into a solution
Context, integration and quality for AI that helps in real processes.
RAG and integrations
Agents connected to documents, APIs, websites and WhatsApp, with domain-specific context.
Quality and traceability
Guardrails, evaluation, sources and non-invention rules reduce unsupported answers.
Hands-on
AI projects
FastAPIChromaDBWhatsApp
Professional AI Framework
Own product in production
Multi-tenant platform for professionals and businesses to create agents connected to a website and WhatsApp, with per-agent RAG, an admin panel and public pages.
Explore the product →
PythonLangChainMCP
HR Policy Agent + MCP
Public demo · sourced RAG
HR policy agent with a versioned corpus, grounded answers, an HTTP API and Model Context Protocol integration.
Open demo →
LangGraphChromaDBRAGAS
ReqGuard
Applied research · in progress
Requirements and compliance analysis system with RAG, a LangGraph flow and quality evaluation.
View on GitHub →
Dialogflow CXVertex AI SearchCloud Run
Aurora Insurance
Portfolio case · Dialogflow CX + Vertex AI Search
Portuguese-language agent that separates document questions with RAG from protocol lookups through a webhook.
Open demo →
FastAPIUser storiesBDD/UAT
AI Requirements Copilot
Requirements prototype
Application to support user stories, BDD, UAT and documentation.
TwilioWhatsAppOpenAI
Healthcare / Twilio POC
Conversational AI
Flow for WhatsApp scheduling and customer service.
SxSFact-checkingRLHF
LLM evaluation
Turing · Outlier AI
SxS evaluation, fact-checking and error analysis for response quality and adherence.
Typebotn8nLLMs
Conversational agents
LIFEE
Personas, flows, rules and integrations using Typebot, n8n and LLM APIs.
Tooling
Working stack
Applied AI
LLMsRAGLangChainLangGraphMCPChromaDB
Integration and operations
PythonFastAPIREST APIsPostgreSQLWhatsAppn8n
Quality and context
Prompt evaluationFact-checkingRAGASGuardrailsRequirementsTraceability
Live demonstrations
An own product that turns context into customer service.
The Professional AI Framework is the main evidence: one product joining business context, service channels and management.