Skip to content
Felipe Iglesias
Contact

Open to opportunities

,,,:Ibuildeverylayerbetweenthequestionandthedecision.

each lit word retrieves a corpus concept · hover or TabPrompt → Prompt engineering · cos 0.382close to response · cos 0.519close to context · cos 0.483context → Personal development platform · cos 0.334close to Prompt · cos 0.483close to response · cos 0.416 · same concept: Personal development platformretrieval → RAG · cos 0.485close to response · cos 0.433response → Personal development platform · cos 0.299close to Prompt · cos 0.519close to retrieval · cos 0.433close to context · cos 0.416 · same concept: Personal development platform

Two years building generative AI that actually ships: RAG on Amazon Bedrock, guardrails, and data pipelines across AWS and GCP.

4 words · 3 corpus concepts · text-embedding-3-small · 512d

View projectsGet in touch

Selected work

01

Contract and public tender analysis

Generative AI and RAG applied to public tenders and contracts.

2024 · Global Hitss
  • python
  • amazon bedrock
  • rag
  • embeddings
  • aws
02

Natural language data querying

A chatbot that writes the SQL: whoever has the question stops depending on whoever can query.

2025 · Global Hitss
  • llm
  • text-to-sql
  • postgresql
  • bigquery
  • gcp
  • guardrails
03

AI powered personal development platform

A commercial product in production: contextual mentor, per area diagnosis, and a 90 day plan the interface renders as an object rather than as text.

2026 · Client product
  • typescript
  • node
  • react native
  • openai api
  • postgresql
04

Hotel management system

A commercial product in production, with the financial module modelled as a ledger: a mistake becomes a reversal, never an UPDATE.

2026 · Client product
  • typescript
  • nestjs
  • prisma
  • postgresql
  • react
  • lgpd
05

ElderCare Watch

A Mamdani fuzzy inference engine written from scratch, in Swift.

Undergraduate thesis · UFF
  • swift
  • fuzzy logic
  • ios

Also

  • Personal iOS projects · Apps in Swift and SwiftUI, built independently.
  • Academic data pipeline · Ingestion, transformation and batch load orchestration.

Ask this portfolio

Ask anything about me. The answer cites its source.

A RAG system over this site’s content. Retrieved passages appear before the answer, with their similarity score, and every citation is checked against them. You can verify where each claim came from. When the site holds no evidence, it says so instead of inventing.

Open the full page

Answers are generated from this site’s content, with mandatory and verified citations.

Experience

  1. Global Hitss

    AI Engineer

    Data and artificial intelligence solutions in a corporate context, with language models in production, RAG on Amazon Bedrock and pipelines on AWS and GCP.

  2. Own projects

    Independent developer

    LLM products delivered end to end for clients, from requirements to production operation. Two of them sold as commercial products.

  3. Universidade Federal Fluminense

    BSc in Computer Science

    Thesis on fuzzy logic applied to health monitoring, with a Mamdani inference engine implemented from scratch.

Stack and certifications

Contact

Open to conversations about AI Engineer roles.

If you are hiring and something here resonated, get in touch. I reply to every message.

Open to opportunities