Carine Toure

Benedict Steen & Associates is a Belgian firm specializing in risk consulting, economic intelligence, and corporate security. With a presence in Europe and Africa and over 25 years of experience in business law, managing over 1000 cases, their mission is to provide privileged access to African, European, American, and Asian markets.

I joined the project team to design a legal information dissemination platform integrating a chatbot specialized in legal matters.

During this assignment, I performed:

- Feasibility study, needs specification, and prototyping

- Python development of data collection and preparation pipelines (web scraping of relevant legal platforms, mapping of legal contracts and extraction of relevant content, data preparation, embedding, and storage)

- Model benchmarking and RAG

- Prompt testing

- Solution architecture design and documentation

Technical context: Python & model development (Langchain, Requests, Transformers, ChatOpenAI), APIs (AWS Opensearch, OpenAI API), Integration (AWS, Postman), Agile practices (Scrum, Github), Python Developer

The University of Lagunes is a private Ivorian higher education and research institution founded by the International Center for Law Development (ICLD). The University of Lagunes has over 300 students across 3 faculties.

I joined the Faculty of Science and Technology as a visiting lecturer responsible for the modules:

- Supervised Machine Learning

- Generalized Linear Models (Statistics)

- and Big Data

for undergraduate Mathematics students and for the Data Science track reserved for professionals.

Skills: Statistics, Linear Models, Big Data Solution Architecture, Apache Spark, PySpark, Map-reduce Framework, Higher Education, Python Developer, Python Development

Cap Collectif is a French leader in civic tech. They develop open-source, French-made tools for consulting citizen communities. Cap Collectif offers 10 consultative tools in SaaS mode, with over 400 public and private decision-maker clients who have trusted them for nearly 10 years.

I joined the AI team for a project analyzing citizen feedback using AI-NLP models for syntactic analysis, sentiment analysis, classification, and text similarity grouping.

During this 2-year assignment, I worked on:

- Project technical specifications

- Language model benchmarking

- Python (Python programming) development of data collection and preparation pipelines

- Model training for various features

- Documentation

- Model deployment based on a Serverless architecture with GraphQL API communication

- Project management according to Scrum principles (backlog definition, task planning and prioritization in coordination with the team lead)

Technical context: Python Development (Pandas, Scikitlearn, Requests, PyTorch, Transformers), NLP Libraries & Models (Nltk, Spacy, TextBlob, BERT, BERTScore), APIs (AWS Boto3, IBM API), Integration (AWS, GraphQL, Openfaas for Kubernetes, Scaleway, GCP, IBM cloud), Agile practices (Scrum, Github)