Back to Portfolio

María Sebares

Data Scientist / AI Engineer

London, United Kingdom
mariasebares9@gmail.com
07561076885

Professional Summary

AI Engineer and Data Scientist with over 3 years of experience specialising in cutting-edge Agentic AI systems and 'full-stack' AI/ML system workflows. Expert in building, deploying, and scaling intelligent Agentic AI solutions using frameworks like DSPy and LangGraph. Adept at leveraging cloud platforms (GCP, AWS) to develop scalable APIs and integrate advanced conversational logic. Proven track record of engaging clients to translate business needs into robust AI-enabled solutions. Committed to advancing the field of AI with a strong foundation in neuroscience and continuous professional growth.

Employment History

Consultant Data Scientist / AI Engineer

Nov 2022 – Present
IBM | London, UK
  • Developed an Agentic Gen-AI Virtual Assistant, significantly improving customer spending insights and driving revenue growth.
  • Led development of complex conversational logic (Multi-turn, context-aware, guardrails) using DSPy, LangGraph, and OpenAI Agents SDK.
  • Implemented production-ready APIs using FastAPI to facilitate real-time system interactions.
  • Managed client communication from POC to full deployment, securing a total contract value of £5M+ over 5 quarters.
  • Developed predictive financial models using XGBoost, SVM, and Random Forests to improve forecasting accuracy.
  • Oversaw version control workflows (GitHub) and enforced team-wide coding best practices.

Intern

Aug 2021
Centre for Finance, Technology and Entrepreneurship (CFTE) | London
  • Conducted research on emerging FinTech trends, focusing on blockchain and AI applications.
  • Assisted in preparing analytical reports on the impact of technology on traditional finance.
  • Supported the development of educational content for online courses and webinars.

Side Projects

Real Estate AI Agent (Tumai)

Jan 2024 – Present
Agentic Document Assistant
  • Designed an AI assistant to help real estate professionals manage documents via natural language chat.
  • Enabled functionality for uploading, retrieving, and emailing documents by integrating backend actions with LLMs.
  • Built with Python, OpenAI Agents SDK, PostgreSQL, and React, utilizing embeddings for intelligent file handling.

Skills

Languages & Core: Machine Learning, Python, SQL, JavaScript, Go, React, Node.js
Agentic Frameworks: LangGraph, DSPy, OpenAI
Cloud & Infra: AWS, GCP, PostgreSQL, Docker, Git
Other: Prompt Engineering, Team Leadership, N8N

Education

University College London (UCL)
MSci Neuroscience (Upper 2.1)
Sep 2018 - Jun 2022
Imperial College London
Mathematics for Machine Learning
Jun 2021 - Aug 2021

Certifications & Awards

  • AWS Cloud Practitioner
  • Datathon Winner 2024, IBM
  • Machine Learning Award, Lloyds

Languages

  • English: Native
  • Spanish: Native
  • French: Highly proficient