Antônio Tameirão

Designed and developed scalable agentic AI systems to support ADP’s global expansion, automating the collection and consolidation of payroll, tax, and compliance data across countries. Built and orchestrated LLM-based agents using Strands and LangGraph, including custom tools, MCPs, and RAG pipelines with MongoDB. Contributed to backend development with FastAPI and deployed production-grade services on AWS using Docker and Kubernetes, leveraging SNS/SQS for event-driven processing. Implemented evaluation (RAGAS), observability (Langfuse), and continuous improvement workflows, ensuring robust testing, monitoring, and maintenance of AI systems in production.

- Designed and continuously improved an AI platform that enables business users to easily access predictive models, data insights, and automation tools, streamlining decision-making across teams. - Developed and deployed multiple LLM-based applications, leveraging technologies like OpenAI, LangChain, and LangGraph to build intelligent, task-specific agents. - Built a production-grade data extraction tool to parse structured information from PDF and Excel files, using OCR and custom parsing logic to automate document processing at scale. - Delivered machine learning Proofs of Concept (POCs) to validate ideas, demonstrate business value, and accelerate the adoption of AI solutions.

- Designed and developed an AI-driven virtual assistant for efficient and humanized debt negotiation over WhatsApp, leveraging OpenAI APIs, LangGraph, LangChain, and LangSmith for advanced natural language processing and agent orchestration. - Built scalable REST APIs and microservices using FastAPI, and managed deployment with Docker to ensure portability and reliability. - Architected and maintained an MSSQL database to store all interactions and settings of the virtual assistant, using SQLAlchemy and Alembic for robust ORM and version-controlled migrations.

Led a team of developers in the ongoing management and enhancement of iOptimum, a cloud-based software solution for maintenance schedule optimization using AI and linear programming. Originally joined the project as the primary developer in my role as a Data Scientist, where I built core optimization features. Over time, transitioned into a leadership role, taking ownership of the strategic roadmap, guiding feature development, and aligning the product with business and client needs.

- Developed and deployed a cloud-based system for maintenance schedule optimization using AI and linear programming. Achieved an average 20% increase in scheduled activities across diverse industries. - Developed and cloud-deployed a deep learning model to forecast key quality metrics in the pulp and paper industry, enhancing planning accuracy and operational efficiency. - Developed and cloud-deployed a machine learning model to identify operational anomalies in coke oven processes, reducing downtime and supporting preventative maintenance. - Leadership and client engagement.

Developed supervised and unsupervised machine learning models to solve railway transportation problems in the mining industry.

Project aimed at participating in the annual SAE BRASIL AeroDesign competition, focusing on developing an aeronautical project through all phases: design, detailing, construction, and testing. Areas of expertise: Aerodynamics and Performance (having been the head of both areas). Achievements: 7th, 14th, and 5th place out of 60 teams (overall ranking) and 2nd best aerodynamics report. - Assisted in developing a multidisciplinary optimization algorithm for aircraft using MATLAB; - Implemented algorithms to evaluate aircraft performance; - Built a test bench for Glow engines, measuring thrust and rotation; - Conducted comparative analysis of theoretical and measured performance during flight tests; - Developed optimization algorithms for wingtip devices; - Conducted CFD simulations of the aircraft for various flight conditions; - Determined the required torque for moving control surfaces; - Performed parametric analyses to predict ground effect; - Assisted in developing the project presentation, achieving 5th place out of 60 teams.

Responsable for running the lifts and keeping it safe for the guests.

Responsable for running the lifts and keeping it safe for the guests.