Carlos Samey

AI/ML engineer with 5+ years designing, building, and deploying LLM-powered applications, RAG systems, and production ML services that reach users at scale across Africa and Asia. Deep hands-on experience with retrieval-augmented generation, recommendation systems, NLP over multilingual and low-resource contexts, and fine-tuning open-weight models. Currently leading a 4-person engineering team on Ask-a-Question, an open-source AI-powered chatbot SaaS, and shipping AI features for partners in global health, public finance, and education.

Comfortable translating requirements from product managers, researchers, and government partners into reliable, observable ML systems.

Data Scientist at IDinsight (2022-09 – Present)

Leads a 4-person engineering team on Ask-a-Question, an open-source AI-powered SaaS chatbot, working end-to-end from product discovery to deployment with product managers, data scientists, and business stakeholders.

• Designed and deployed an LLM-powered healthcare chatbot with RAG-based document retrieval to help pregnant women in South Africa access reliable health information; added chat and voice channels and built data pipelines for document ingestion and retrieval.
• Scoped and built GenAI feedback analytics for Public Finance Management in partnership with the Government of Kenya, including ETL pipelines that turn unstructured text into decision-ready insights.
• Built an AI agent supporting bilingual education in partnership with the Government of Senegal, applying multilingual text processing and NLP techniques.

Machine Learning Engineer at Gozem (2021-11 – 2022-08)

• Shipped a personalized recommendation system for food delivery, partnering with the marketing team; measurable lift in user engagement and retention inside a live product.
• Implemented an improved distance-calculation algorithm inside the Gozem product, increasing accuracy by up to 40%.

Data Scientist & Consultant at Togolese Ministry of Digital Economy (2021-05 – 2021-11)

• Led scoping of AI/OCR models to streamline government data collection, in partnership with the Ministry of Tourism.

Data Science Intern at Equinox AI (2021-05 – 2021-11)

• Built deep-learning models to forecast solar energy production, reaching 77% accuracy.
• Optimized ETL pipelines and automated model deployment in production environments.

MSc in Data Science – Middlesex University of London

BSc in Mathematics and Computer Science – Catholic University of West Africa