Gustavo is available for hire
Hire GustavoGustavo Franco
Verified Expert in Engineering
AI Architect, Lead Data Scientist, and Developer
Providence, RI, United States
Toptal member since March 19, 2026
Bio
Gustavo is a versatile AI architect, lead data scientist, and prompt engineer with 12+ years of experience. He has designed and delivered scalable, production-grade machine learning (ML) and generative AI (GenAI) solutions for government, retail, and enterprise clients. Gustavo is an expert in large language models (LLMs), agentic systems, NLP, MLOps, and cloud-based infrastructure, including Azure and AWS.
Portfolio
ICF International
Python, Databricks, Large Language Models (LLMs), DevOps...
Atrium
Machine Learning Operations (MLOps), MLflow, Machine Learning...
VMWare
Python, Machine Learning Operations (MLOps), MLflow, Spark, PySpark, Dataiku...
Experience
- Machine Learning - 12 years
- DevOps - 12 years
- Python - 12 years
- Natural Language Processing (NLP) - 11 years
- Databricks - 7 years
- Large Language Models (LLMs) - 6 years
- LangChain - 6 years
- RAG Architecture - 4 years
Preferred Environment
Python, LangChain, Databricks, Dataiku, FastAPI, MLflow, RAG Architecture, Large Language Models (LLMs), Machine Learning, Natural Language Processing (NLP)
The most amazing...
...system I've built is an end-to-end agentic AI platform that transforms raw, unstructured feedback and structured data into actionable, self-service insights.
Work Experience
Senior AI Engineer
2024 - 2026
ICF International
- Architected a production-grade multi-agent LLM platform using Llama 3 (70B) that transformed unstructured customer feedback into actionable insights through sentiment analysis, categorization, and executive summarization.
- Conducted experiments and developed POCs demonstrating the impact of LLM agents and RAG to address key voice of customer (VoC) challenges in online passport services, increasing stakeholder buy-in across departments.
- Designed natural language interfaces that enabled users to query structured survey data using AI agents capable of reformulating requests, retrieving data, and generating contextual responses.
- Implemented prompt engineering, evaluation frameworks, and statistical validation that achieved human-level sentiment classification accuracy at a 95% confidence interval.
- Established CI/CD pipelines, automated testing, prompt versioning, and reusable Python libraries that improved reliability, maintainability, and deployment consistency across AI workflows.
- Developed REST APIs, output validation frameworks, and monitoring pipelines that enabled secure integration of LLM capabilities into enterprise dashboards and downstream applications.
- Mentored junior engineers and advised technical leadership on AI architecture, model governance, and scalable deployment strategies for production generative AI systems.
Technologies: Python, Databricks, Large Language Models (LLMs), DevOps, Natural Language Processing (NLP), Data Science, ETL, Azure Databricks, Data Engineering, CI/CD Pipelines, Apache Spark, PySpark, Machine Learning Operations (MLOps), Data Analysis, Data Pipelines, PostgreSQL, SQL, Docker, Artificial Intelligence (AI), APIs, Azure, AI Integration, Git, Data Integration, AI Development, AI Architecture, RAG Architecture, Generative Artificial Intelligence (GenAI), Agentic AI, Prompt Engineering, AI Agents, Multi-agent Orchestration, Model Context Protocol (MCP), LLM Agents, Delta Lake, Unity Catalog, FastAPI, LangChain, Machine Learning, TypeScript, Amazon S3 (AWS S3), Linux, AI Pipeline, Metrics, Data Mining, Unit Testing, Sentiment Analysis, Spark NLP, Software Development, Testing, Refactoring, Agentic RAG Systems, RAG Systems, Retrieval-augmented Generation (RAG), Azure OpenAI Service, Automation, Data Processing, Research, LangGraph, Back-end, Agentic Frameworks, Terraform, Large Language Model Operations (LLMOps), Agentic AI Systems, Architecture, Agent Deployment, Cloud, Cloud Services, Google Cloud Platform, Technical Documentation, Data Analytics
Lead Data Scientist
2023 - 2024
Atrium
- Architected and optimized a customer retention model using NLP and GenAI techniques, increasing ROI from high-value repeat customers by over 15%.
- Led the project management for customer retention strategies and customer-facing initiatives, ensuring timely delivery and stakeholder satisfaction.
- Explored advanced LLM-based text-to-SQL solutions, aiming to improve data accessibility for non-technical users.
- Developed LLM-powered SQL-to-Text and analytics explanation systems using few-shot prompting and chain-of-thought reasoning, improving accessibility of enterprise data.
- Migrated legacy machine learning workflows to Dataiku, improving scalability, deployment consistency, and operational efficiency across production pipelines.
- Implemented agent-based workflows with tool calling and automated data interpretation, enabling natural language interaction with enterprise datasets.
- Established monitoring, alerting, unit testing, and CI/CD practices that improved production reliability and reduced operational downtime for machine learning systems.
- Led cross-functional AI initiatives while mentoring junior data scientists and delivering customer-facing machine learning solutions from concept through production deployment.
Technologies: Machine Learning Operations (MLOps), MLflow, Machine Learning, Natural Language Processing (NLP), Large Language Models (LLMs), Python, Data Science, ETL, Data Engineering, CI/CD Pipelines, Apache Spark, PySpark, Data Analysis, Data Pipelines, PostgreSQL, SQL, Amazon Web Services (AWS), Docker, Artificial Intelligence (AI), APIs, Google Cloud Platform (GCP), Amazon S3 (AWS S3), Snowflake, AI Integration, Git, Data Integration, AI Development, AI Architecture, LLM Agents, LLM Integration, Generative Artificial Intelligence (GenAI), AI Agents, Agentic AI, RAG Architecture, RAG Pipelines, Multi-agent Orchestration, Model Context Protocol (MCP), Prompt Engineering, FastAPI, Vector Databases, Pytest, Model Monitoring, Word Embedding, LangChain, Databricks, DevOps, TypeScript, Azure Databricks, Linux, AI Pipeline, Metrics, Unit Testing, Sentiment Analysis, Spark NLP, Software Development, Testing, Refactoring, Agentic RAG Systems, RAG Systems, Retrieval-augmented Generation (RAG), Automation, Data Processing, Research, LangGraph, Back-end, Agentic Frameworks, Terraform, Large Language Model Operations (LLMOps), Agentic AI Systems, Architecture, Cortex, SnowFlake, Agent Deployment, Cloud, Cloud Services, Technical Documentation, Data Analytics
ML Specialist
2022 - 2023
VMWare
- Developed customer recommendation and acquisition models using machine learning and customer segmentation, increasing acquisition rates by 18% and improving customer retention.
- Built a customer acquisition model using advanced ML techniques, resulting in a 22% increase in customer retention.
- Built production-ready machine learning pipelines in Dataiku with integrated logging, testing, and monitoring to improve model reliability and operational scalability.
- Created reusable Python libraries and engineering utilities that standardized data access, error handling, and logging across multiple AI development teams.
- Implemented monitoring and KPI alerting systems that reduced troubleshooting time and improved visibility into production model performance.
- Collaborated with engineering and business stakeholders to define analytics metrics, optimize recommendation systems, and support enterprise AI initiatives.
- Standardized Python development practices by authoring coding guidelines and improving software quality, maintainability, and engineering productivity across the team.
Technologies: Python, Machine Learning Operations (MLOps), MLflow, Spark, PySpark, Dataiku, Data Science, Data Engineering, CI/CD Pipelines, Apache Spark, Data Analysis, Data Pipelines, PostgreSQL, SQL, Amazon Web Services (AWS), Docker, Artificial Intelligence (AI), APIs, Google Cloud Platform (GCP), Amazon S3 (AWS S3), AI Integration, Git, Data Integration, Time Series, AI Development, AI Architecture, Databricks, Machine Learning, DevOps, TypeScript, ETL, Linux, Data Modeling, AI Pipeline, Metrics, Data Mining, Unit Testing, Sentiment Analysis, Spark NLP, Software Development, Testing, Refactoring, Agentic RAG Systems, RAG Systems, Retrieval-augmented Generation (RAG), Automation, Data Processing, Research, LangGraph, Back-end, Agentic Frameworks, Terraform, Geospatial Data, Large Language Model Operations (LLMOps), Agentic AI Systems, Architecture, Cloud, Cloud Services, Technical Documentation, Data Analytics
Senior Data Scientist
2019 - 2021
Deloitte
- Architected a POC that automated complex tax impact calculations for financial institutions, simplifying regulatory analysis and demonstrating the feasibility of applying machine learning and advanced analytics to enterprise tax modeling.
- Led a cross-functional team of 4 engineers and data scientists, providing technical leadership, mentoring, and coordinating project execution while collaborating closely with business stakeholders to align technical deliverables with the objectives.
- Developed ML and clustering models to identify bot activity during high-demand product launches, delivering actionable insights that improved launch fairness, enhanced user experience, and strengthened the integrity of Nike's digital platform.
- Designed and implemented validation frameworks for Themis geolocation models while analyzing large-scale user behavior data to improve bot detection accuracy and support fair access to limited-release product launches at Nike.
- Architected scalable data pipelines that monitored infrastructure database compliance and operational metrics, improving visibility across enterprise systems at Facebook.
- Developed executive dashboards that visualized database compliance and AppWeight metrics, enabling Facebook engineering leadership to make faster, data-driven decisions.
Technologies: Machine Learning Operations (MLOps), Machine Learning, Data Science, Data Engineering, Dataiku, ETL, DevOps, AI Development, Artificial Intelligence (AI), Git, Linux, DataRobot, Data Pipelines, Data Modeling, CI/CD Pipelines, AI Pipeline, Metrics, AI Integration, Quantitative Analysis, Natural Language Processing (NLP), TypeScript, Data Analysis, PostgreSQL, Amazon S3 (AWS S3), Unit Testing, Sentiment Analysis, Spark NLP, Software Development, Testing, Refactoring, Automation, Data Processing, Research, Back-end, Agentic Frameworks, Terraform, Geospatial Data, Architecture, Cloud, Google Cloud Platform
Senior Data Scientist
2018 - 2019
athenahealth
- Developed machine learning and NLP solutions for entity resolution and patient matching, achieving 80% accuracy in identifying duplicate patient records without unique identifiers, improving healthcare data quality and patient record integrity.
- Architected named-entity recognition (NER) models using custom Word2Vec embeddings and deep learning techniques, enabling accurate extraction of clinical entities from unstructured healthcare text.
- Designed proofs-of-concept evaluating Redis, Elasticsearch, and AWS storage services, identifying cost-effective database architectures that improved scalability and performance for enterprise applications.
- Built production-ready forecasting tools in Python and Flask, enabling business users to deploy forecasting models through interactive dashboards with configurable prediction workflows.
- Implemented cloud-native machine learning workflows using AWS (EC2, S3, CloudWatch), Docker, Jenkins, and Git, improving deployment automation, CI/CD reliability, and operational efficiency.
- Developed predictive models using convolutional neural networks, K-Nearest Neighbors, and statistical learning techniques to support healthcare analytics, entity matching, and product intelligence initiatives.
Technologies: AI Architecture, Amazon S3 (AWS S3), APIs, Amazon Web Services (AWS), Artificial Intelligence (AI), AI Pipeline, Data Analysis, AI Development, AI Integration, Quantitative Analysis, PySpark, DevOps, Docker, Keras, Natural Language Processing (NLP), NLU, Optical Character Recognition (OCR), Spark, Hadoop, CI/CD Pipelines, Jenkins, Unit Testing, SQL, Redis, Caché, PostgreSQL, Machine Learning, Machine Learning Operations (MLOps), MLflow, Named-entity Recognition (NER), Word2Vec, Word Embedding, TypeScript, ETL, Data Science, Data Pipelines, Linux, Data Modeling, Metrics, Data Mining, Software Development, Testing, Refactoring, Automation, Data Processing, Research, Back-end, Terraform, Architecture
Data Scientist II
2016 - 2018
Johnson & Johnson
- Led the design and development of an end-to-end sales forecasting platform, leveraging machine learning and statistical modeling to improve demand planning, inventory forecasting, and business decision-making across multiple product lines.
- Developed predictive machine learning models to analyze customer behavior, product quality, and sales performance, delivering actionable insights that improved product monitoring and supported data-driven commercial strategies.
- Architected a global counterfeit detection solution using NLP, Latent Dirichlet Allocation (LDA), clustering algorithms, and anomaly detection to identify counterfeit products and uncover geographic distribution patterns across international markets.
- Implemented natural language processing and sentiment analysis pipelines that mined social media data to measure customer sentiment and competitive product perception, enabling product teams to better understand market trends and behavior.
- Optimized supply chain operations by developing Python-based optimization models, modernizing CI/CD pipelines with Jenkins, and building interactive Tableau dashboards that improved operational visibility and executive reporting.
- Led the development of a global counterfeit detection platform using NLP, clustering, and anomaly detection to identify fraudulent products and uncover international distribution patterns.
Technologies: Linear Discriminant Analysis (LDA), Amazon S3 (AWS S3), R, Natural Language Processing (NLP), Data Mining, Git, Jenkins, Jira, Tableau, Time Series, PySpark, Keras, Continuous Delivery (CD), Amazon Web Services (AWS), SQL, PostgreSQL, Unit Testing, Random Forests, APIs, Revenue Projections, Machine Learning, DevOps, TypeScript, Machine Learning Operations (MLOps), ETL, Data Science, Data Analysis, Data Pipelines, Linux, Data Modeling, Metrics, Software Development, Testing, Refactoring, Automation, Data Processing, Research, Geospatial Data
Experience
Voice of the Customer AI Platform
At ICF, I designed and deployed a multi-agent LLM system (Llama 3 70B) that allowed users to interact with complex datasets using natural language. Instead of relying on dashboards or analysts, users could ask questions, and the system would:
• Reformulate the query intelligently (handling ambiguity and intent)
• Route the request to the right tools or data sources
• Query structured databases and analytics pipelines
• Interpret the results contextually
• Generate a clear, business-ready answer
At the same time, the system processed tens of thousands of open-ended feedback entries through a three-stage pipeline (sentiment, themes, and insights), enabling leadership to understand customer pain points quickly.
• Reformulate the query intelligently (handling ambiguity and intent)
• Route the request to the right tools or data sources
• Query structured databases and analytics pipelines
• Interpret the results contextually
• Generate a clear, business-ready answer
At the same time, the system processed tens of thousands of open-ended feedback entries through a three-stage pipeline (sentiment, themes, and insights), enabling leadership to understand customer pain points quickly.
Education
2019 - 2026
PhD in Informatics and Applied Mathematics
University of Massachusetts Dartmouth - Dartmouth, MA, USA
2019 - 2022
Master's Degree in Data Science
University of Massachusetts Dartmouth - Dartmouth, MA, USA
2011 - 2017
Bachelor's Degree in Mathematics and Computer Science
University of Massachusetts Dartmouth - Dartmouth, MA, USA
Skills
Libraries/APIs
PySpark, Pandas, Spark NLP, D3.js, Keras
Tools
Git, Claude, Codex, Claude Code, Azure OpenAI Service, Terraform, Tableau, Microsoft Power BI, DataRobot, Jenkins, Jira, Named-entity Recognition (NER), Pytest
Languages
Python, TypeScript, SQL, Snowflake, R
Frameworks
Spark, Apache Spark, LangGraph, Agentic Frameworks, Hadoop
Paradigms
DevOps, ETL, Testing, Refactoring, Automation, Model Context Protocol (MCP), Continuous Delivery (CD), Unit Testing
Platforms
Databricks, Amazon Web Services (AWS), Azure, Cortex, Dataiku, Docker, Google Cloud Platform (GCP), Kubernetes, Vertex AI, Linux
Storage
Data Pipelines, PostgreSQL, Amazon S3 (AWS S3), Data Integration, Redis, Caché
Other
LangChain, RAG Architecture, Large Language Models (LLMs), Machine Learning, Natural Language Processing (NLP), Machine Learning Operations (MLOps), CI/CD Pipelines, Data Engineering, Azure Databricks, Data Science, Data Analysis, Artificial Intelligence (AI), APIs, AI Integration, Time Series, AI Development, AI Architecture, Data Modeling, AI Pipeline, Metrics, Data Mining, Word2Vec, Word Embedding, Generative Artificial Intelligence (GenAI), Agentic AI, Prompt Engineering, AI Agents, Multi-agent Orchestration, LLM Agents, LLM Integration, RAG Pipelines, Software Architecture, Distributed Systems, Stakeholder Management, Team Mentoring, Semantic Search, Gemini, AI Agent Orchestration, Error Handling, AI Modeling, AI Model Training, AI Engineering, Sentiment Analysis, Software Development, Agentic RAG Systems, RAG Systems, Retrieval-augmented Generation (RAG), Data Processing, Research, Back-end, Geospatial Data, Large Language Model Operations (LLMOps), Agentic AI Systems, Architecture, Agent Deployment, SnowFlake, Cloud, Cloud Services, Google Cloud Platform, Technical Documentation, Data Analytics, Applied Mathematics, Mathematics, Statistics, MLflow, Cloud Architecture, Technical Leadership, Enterprise Architecture, System Design, FastAPI, Quantitative Analysis, Linear Discriminant Analysis (LDA), Random Forests, Revenue Projections, NLU, Optical Character Recognition (OCR), Delta Lake, Unity Catalog, Vector Databases, Model Monitoring
Collaboration That Works
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
Start your risk-free talent trial
Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.
Top talent is in high demand.
Start hiring