marcos padilla

Technical Lead - AI Engineering

Armada

Jul 2025 - Current (1 year 3 months)

Led architecture and delivery of production-grade LLM and multimodal AI systems, reducing end-to-end solution delivery time by 40% through standardized model pipelines and reusable components. Defined Al technical strategy across multiple initiatives, improving cross-team execution velocity by 30% and reducing rework caused by unclear requirements by 25%. Mentored and technically guided senior engineers, increasing team-level code quality and model reproducibility, resulting in 35% fewer post-deployment issues. Introduced best practices for model evaluation, safety, and monitoring, improving model reliability and reducing inference failures by 45% in production. Integrated cutting-edge research (RAG, multimodal fusion, RLHF-inspired fine-tu

Principal AI Engineer

Armada

Jun 2024 - Jul 2025 (1 year 2 months)

Designed and fine-tuned LLM and generative AI pipelines using Azure AI Studio and Azure OpenAI, leveraging prompt flows, evaluation pipelines, and grounded generation to achieve 25-40% improvements in task accuracy, response relevance, and hallucination reduction. Built and operationalized end-to-end GenAI workflows using Azure AI Foundry, enabling standardized model lifecycle management, secure access control, and governed deployment of LLM-powered applications across teams. Implemented retrieval-augmented generation (RAG) architectures in Azure AI Studio, integrating Azure OpenAI with Azure Blob Storage and Azure Cognitive Search to ground model responses in enterprise knowledge. Established scalable MLOps infrastructure on Azure (trainin

Senior AI Engineer

Armada

Jul 2023 - Jun 2024 (1 year)

Developed and deployed deep learning and NLP models that improved core system accuracy and automation rates by 20-25%. Implemented end-to-end ML pipelines (data ingestion → training → deployment), reducing manual intervention and operational overhead by 40%. Partnered with product teams to deliver AI-driven features that reduced customer processing time by 30% and improved user satisfaction metrics. Improved model inference latency by 25% through optimization of architectures, preprocessing, and runtime execution. Contributed to internal tooling and open-source initiatives, accelerating team onboarding and reducing ramp-up time for new engineers by 20%.

Data Scientist

American Airlines

Aug 2022 - Jul 2023 (1 year)

Led the development of a deep neural network-based forecasting engine to predict flight-level traffic demand, supporting data-driven planning and operational decision-making. Designed and implemented a multimodal neural network architecture that combined historical demand signals with multi-dimensional operational data, improving forecast accuracy compared to traditional baselines. Evaluated model performance across multiple flight routes and market conditions to assess robustness and adaptability to shifting demand trends. Guided and mentored a team of four data scientists and engineers, coordinating experimentation, model refinement, and validation workflows. Collaborated with stakeholders to translate forecasting results into actionable

Machine Learning Postdoctoral Researcher

Truveta

May 2022 - Aug 2022 (4 months)

Worked on data-driven NLP approaches to support the company's vision of Saving Lives with Data. Pre-trained and Fine-tuned Large Language Models (LLM) on clinical text data, for up-stream and down-stream tasks.

R

Data Scientist Research

Redshred

Mar 2021 - May 2022 (1 year 3 months)

Implemented a novel plot-processing approach utilizing object detection models and text extraction techniques, resulting in a 69.12% success rate in correctly identifying data from plots, approaching the state-of-the-art in the field. Deployed on the RedShred API for automatic processing of PDF documents. Utilized HuggingFace framework to develop and train vision transformer-based methods (ViT) for document understanding, including segmentation and object detection. These techniques demonstrated strong performance in the field of document analysis.

P

Software Engineer

PMAD

May 2019 - May 2021 (2 years 1 month)

Led the end-to-end development of two e-commerce platforms, using Ruby on Rails, WordPress, Next.js, and Shopify, delivering fast, secure, and scalable online storefronts. Designed and implemented custom themes, plugins, and integrations to support advanced product catalogs, checkout flows, and customer experiences. Optimized site performance, SEO, and Core Web Vitals, improving load times and overall usability. Collaborated with cross-functional teams to gather requirements, plan releases, and ensure smooth deployment and maintenance of production systems. Implemented best practices in code quality, testing, and CI/CD to support long-term maintainability of the platforms. Contributed to backend development and web applications using Ruby o

About

Detail

Timeline

Résumé

marcos padilla

About

Detail

Behavioral traits