Work

I build production AI systems that improve cost, quality, and throughput across banking, startups, and open-source software.

Case studies

C6 Bank

Multi-agent sales assistant for a 30M+ customer bank

Cut customer acquisition cost by 6x while improving sales team productivity.

The team needed a production AI assistant that could guide sales flows reliably across tools, stakeholders, and business constraints.

Approach

  • Led the end-to-end build of a generative AI sales assistant using LangChain and LangGraph.
  • Applied tool use, personas, workflows, guardrails, and model selection strategies to improve reliability.
  • Worked across backend, MLOps, data, sales, and marketing teams to ship the system.

Stack

LangGraph · LangChain · Python · LLM orchestration · Production evaluation

C6 Bank

RAG optimization for lower cost and better answers

Reduced inference cost by 90% while improving answer quality beyond competing LLM alternatives.

The existing RAG system was too expensive and needed stronger answer quality for a production banking context.

Approach

  • Researched new GCP capabilities and prompt strategies.
  • Reworked retrieval and prompting for stronger accuracy and cost efficiency.
  • Focused on production constraints rather than benchmark-only gains.

Stack

GCP · RAG · Prompt engineering · Evaluation

Qive

Product retrieval redesign for startup search

Cut indexing time from hours to minutes and reached 90%+ recall with only 20 candidates instead of 100.

The product search engine needed materially better retrieval efficiency and fresher indexing over proprietary catalog data.

Approach

  • Analyzed product groupings to redesign indexing strategy.
  • Improved vector retrieval with Gemini-based data augmentation.
  • Reported directly to founders and mentored developers while shipping improvements.

Stack

GCP · Vector retrieval · Gemini · RAG

Career

Senior AI Engineer

Xseed Solutions

Sep 2025 - Present

United States

  • Building and shipping production GenAI systems for U.S.-based clients.
  • Designing retrieval, orchestration, and evaluation pipelines for applied AI products.
  • Collaborating remotely with cross-functional teams on end-to-end AI delivery.

Senior AI Engineer

Qive

Feb 2025 - Aug 2025

Brazil

  • Led development of a product search engine over proprietary data in a founder-led startup.
  • Reduced RAG indexing time from hours to minutes through data-driven grouping and indexing strategies.
  • Achieved 90%+ retrieval recall with 20 candidates instead of 100 using redesigned vector retrieval and Gemini augmentation.
  • Conducted technical interviews and mentored developers.

Senior AI Engineer / Data Scientist

C6 Bank

Oct 2023 - Feb 2025

Brazil

  • Built a multi-agent generative AI sales assistant that reduced acquisition cost by 6x.
  • Improved a production RAG system to cut inference cost by 90% and raise answer quality.
  • Worked across backend, MLOps, data science, sales, and marketing in a 30M+ customer digital bank.
  • Developed forecasting and clustering models that improved customer segmentation and campaign targeting.

Data Science Intern

Argilla

Nov 2021 - May 2022

Spain

  • Built the first production-ready embedding-based annotation methods in the Argilla open-source library.
  • Translated current NLP research into robust product features and supporting tutorials.

AI Engineer

CEXIA / Deep Learning Brasil

Sep 2020 - Jul 2021

Brazil

  • Delivered enterprise NLP proofs of concept across intent detection, NER, sentiment analysis, entity linking, and early generative chatbots.
  • Won first place in ASSIN 2 and ABSAPT with transformer-based systems.

Data Scientist

NeuralMind

Feb 2021 - May 2021

Brazil

  • Won COLIEE 2021 Task 2 with a zero-shot legal entailment approach.

Strengths

LLMs RAG Graph RAG Multi-agent systems Tool use Prompt engineering Evaluation LangGraph LangChain LlamaIndex OpenAI API Anthropic API AWS Bedrock n8n Python PyTorch TensorFlow spaCy Hugging Face BigQuery Apache Spark SQL Vector databases AWS GCP Azure Docker Kubernetes Kubeflow MLflow FastAPI