Tech Lead

Leading the architecture and delivery of production GenAI systems across enterprise use cases, spanning agent-based workflows, LLM orchestration, document intelligence, RAG pipelines, and enterprise integration components.

• Design multi-stage systems combining OCR, structured data extraction, retrieval-augmented reasoning, and human-in-the-loop validation to transform unstructured documents into structured outputs for proposal analysis, decision support, and contract drafting automation.
• Act as Forward Deployed Engineer (FDE), defining end-to-end system architecture, deployment and integration strategies, and adoption paths, while driving integration, rollout, and platform expansion within the client environment.

GenAI Engineer – GenOS AI Team

Contributing to the development of GenOS, DareData’s enterprise AI operating system for deploying, orchestrating, monitoring, and governing AI applications at scale. Building core AI platform capabilities for AI agents, workflow automation, evaluation and observability systems, knowledge management, enterprise integrations, and multi-LLM orchestration, with a focus on production-grade abstractions for GenAI applications through reusable, composable components.

AI Engineer & Data Scientist

Implementing and delivering end-to-end AI, GenAI, and data-driven solutions across telecommunications, energy, public sector, and shared services environments, translating business requirements into production-grade systems.

Selected projects:

• NOS SGPS – Contributed to a company-wide GenAI transformation initiative as part of the SCAILE Program, designing and implementing reusable enterprise AI patterns, including AI agents, workflow automation, knowledge access, and internal copilots, in collaboration with cross-functional teams.
• DREM & Statistics Portugal – Developed a GenAI-powered multi-agent assistant for natural language querying of official statistical data (dynamic SQL generation) and institutional data (multilingual RAG) with strict factual accuracy. Built the supporting ETL pipelines for the Data Warehouse and the APIs powering chatbot and website data access.
• Greenvolt – Improved and maintained a self-service energy community onboarding platform where users upload invoices to obtain savings estimates, customer qualification, and community matching. Built multi-agent workflows for document extraction, validation, and decisioning, integrated with contract generation and DocuSign-based signing flows.
• Porto de Lisboa – Automating an end-to-end public procurement workflow through document information extraction, OCR, inconsistency detection, legal validation, and human-in-the-loop review, ensuring maximum consistency, traceability, and compliance across the process. Beyond the technical gains, the system meaningfully improves the day-to-day quality of life of the people involved, freeing them from repetitive, manual tasks and from drowning in paperwork so they can focus on higher-value decisions.

Key Technologies: OpenAI · Anthropic · LlamaIndex · Google ADK · LangChain · RAG · MCP · Agent2Agent (A2A) · OpenTelemetry · FastAPI · GCP · Docker · Azure · MongoDB · PostgreSQL · Streamlit · etc.

• Design and evaluate complex software engineering tasks for state-of-the-art Grok coding and reasoning models, assessing capabilities across software architecture, debugging, code generation, testing, refactoring, and system design.
• Analyze model performance, robustness, and failure modes in realistic production-like engineering scenarios.
• Contribute to benchmark development, evaluation methodologies, and training signals that improve model reliability, coding quality, and alignment.
• Provide structured insights and expert feedback in collaboration with the technical teams to improve coding performance and overall product quality.