Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I'm Maria Azevedo, Principal AI Engineer focused on LLM-powered systems. I take models from experimentation to production-grade applications, designing RAG pipelines, scalable APIs, and real-time AI workflows that help teams safely act on AI-generated insights. I have a strong background building reliable full-stack systems that translate complex AI behavior into high-impact products, and I collaborate across product, data science, and platform teams to deliver measurable value.…Hi, I'm Maria Azevedo, Principal AI Engineer focused on LLM-powered systems. I take models from experimentation to production-grade applications, designing RAG pipelines, scalable APIs, and real-time AI workflows that help teams safely act on AI-generated insights. I have a strong background building reliable full-stack systems that translate complex AI behavior into high-impact products, and I collaborate across product, data science, and platform teams to deliver measurable value.

Maria Azevedo

AI Engineer, Data Scientist, Full Stack Developer, +5





Hi, I'm Maria Azevedo, Principal AI Engineer focused on LLM-powered systems. I take models from experimentation to production-grade applications, designing RAG pipelines, scalable APIs, and real-time AI workflows that help teams safely act on AI-generated insights. I have a strong background building reliable full-stack systems that translate complex AI behavior into high-impact products, and I collaborate across product, data science, and platform teams to deliver measurable value.…Hi, I'm Maria Azevedo, Principal AI Engineer focused on LLM-powered systems. I take models from experimentation to production-grade applications, designing RAG pipelines, scalable APIs, and real-time AI workflows that help teams safely act on AI-generated insights. I have a strong background building reliable full-stack systems that translate complex AI behavior into high-impact products, and I collaborate across product, data science, and platform teams to deliver measurable value.

Available to hire

Hi, I’m Maria Azevedo, Principal AI Engineer focused on LLM-powered systems. I take models from experimentation to production-grade applications, designing RAG pipelines, scalable APIs, and real-time AI workflows that help teams safely act on AI-generated insights.

I have a strong background building reliable full-stack systems that translate complex AI behavior into high-impact products, and I collaborate across product, data science, and platform teams to deliver measurable value.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Work Experience

Principal AI Engineer at Zest AI

December 1, 2023 - Present

Led end-to-end delivery of LLM-powered features across enterprise systems and user-facing platforms, driving operational efficiency and automation. Designed and implemented Retrieval-Augmented Generation (RAG) pipelines using LangChain and pgvector, enabling internal knowledge Q&A and improved semantic support. Built LLM orchestration and agent APIs with FastAPI, supporting streaming inference, prompt validation, and fallback logic to improve responsiveness and production stability. Developed embedding, hybrid search (BM25 + dense), and reranking microservices to enhance search relevance and recall. Integrated LLM observability and evaluation (LangSmith) to monitor latency, hallucinations, and response quality, increasing governance readiness. Delivered React + TypeScript frontend integrations for LLM-driven assistants and mentored engineers on agent patterns, embedding workflows, and AI safety.

Senior AI Engineer at Apple Inc

August 1, 2020 - December 1, 2023

Designed and shipped AI- and later LLM-powered internal applications, enabling natural-language access to large datasets and reducing reliance on manual data exploration. Built AI application-layer services and backend APIs to orchestrate prompts, retrieval, and responses, improving integration stability and production readiness. Implemented context-aware retrieval pipelines that enriched LLM responses with internal knowledge, improving answer relevance and factual accuracy. Collaborated with data, platform, and product teams to define AI use cases and deliver iterative releases, driving steady adoption of AI-powered workflows. Established AI reliability, observability, and traceability practices, ensuring outputs were auditable, trustworthy, and suitable for enterprise use.

Senior AI Engineer / Lead Machine Learning Engineer at Microsoft

January 1, 2016 - August 1, 2020

Designed and deployed machine learning systems for search relevance, recommendations, and intelligent automation across cloud and internal platforms. Built end-to-end ML pipelines including data ingestion, feature engineering, model training, evaluation, and deployment at scale. Introduced early deep learning models (NLP and ranking) into production workflows as frameworks like TensorFlow matured. Partnered with product teams to translate business problems into data-driven and ML-backed solutions. Served as technical lead on ML initiatives, guiding design reviews and production readiness for multiple teams.

Senior Software Engineer (Data & ML Systems) at Microsoft

July 1, 2010 - December 1, 2015

Developed data-intensive backend services supporting analytics, experimentation, and early machine learning use cases. Built large-scale data processing pipelines using SQL, distributed systems, and early big-data platforms. Implemented predictive models and statistical methods to improve personalization, forecasting, and operational decision-making. Collaborated with data scientists and researchers as the Data Scientist role began to formalize, supporting model integration into production systems.

Web Developer at BX3 Interactive

November 1, 2007 - June 1, 2010

Developed and maintained client-facing web applications using HTML, CSS, JavaScript, and jQuery; implemented PHP/MySQL backend functionality for forms, data processing, and basic business logic. Integrated third-party services to improve reliability, resolved cross-browser compatibility issues, and collaborated with designers to translate Photoshop mockups into production-ready pages, supporting deployments via FTP/SFTP.