Yohan Goncalves

As Tech Lead at Bpifrance, the French public investment bank, dedicated to supporting companies by providing financing, guidance, and investment solutions to foster innovation and growth.

Our team was organized with 6 people: 4 developers, 1 Product Owner, and 1 Scrum master.

Within this team, I designed and deployed a conversational AI platform. This platform, centered around an intelligent chatbot, offers centralized access to a variety of language models through a dedicated proxy. By using RAG and function calling techniques, the chatbot can draw from a vast internal and external knowledge base to provide precise and contextual answers to employees.

Chatbot ~1000 daily users

Proxy ~30 consuming teams

My responsibilities:

- Building the project architecture

- Developing the chatbot (Chainlit, Python)

- Developing the RAG API (FastAPI, Python)

- Developing the agents (LangChain, LangGraph, FastAPI)

- Developing CI/CD pipelines (Jenkinsfile, Jenkins, GitLab)

- Unit testing (Pytest, RAGAS)

- Ensuring technical quality of deliverables (Pair review, SonarQube)

- Production deployment (Helm, Kubernetes)

- Implementing observability (Datadog)

Technical context: Python, FastAPI, PostgreSQL, Kafka, Langfuse, Langchain, LangGraph, RAG, PyTest, RAGAS, Jenkins, Gitlab, AWS, Azure, Kubernetes, Helm, Flux, Datadog

As a DevOps Engineer, I worked at Bpifrance, the French public investment bank, dedicated to supporting companies by providing financing, guidance, and investment solutions to foster innovation and growth.

My role was within a team of 8 people, consisting of 6 DevOps engineers, 1 Product Owner, and 1 Scrum Master.

In this context, I contributed to the design and deployment of DevOps solutions, with a particular focus on improving development and continuous integration practices. One of my major achievements was the design and development of an application for measuring DORA (DevOps Research and Assessment) metrics, aimed at tracking the adoption of DevOps principles within different project teams.

My responsibilities included:

- Assisting teams in adopting and implementing DevOps practices

- Optimizing the Git workflow (Git, GitLab)

- Designing and implementing CI/CD pipelines, accelerating the delivery cycle and enhancing code quality (Jenkinsfile, Jenkins, GitLab, SonarQube)

- Developing an application to measure DORA metrics to monitor DevOps adoption (Java, AWS Lambda, MongoDB)

- Unit testing (JUnit)

- Production deployment (Helm, Kubernetes)

- Implementing observability (Datadog)

Technical context: Java, PostgreSQL, MongoDB, AWS, Kafka, Kubernetes, Helm, Flux, Jenkins, GitLab, Terraform, Atlantis

As Tech Lead at Mind7 Consulting for Colissimo, within a team of 8 people including 1 Scrum Master.

My primary role was the design and development of Scala applications, particularly for consuming and producing Kafka messages, related to parcel delivery. I was also responsible for developing unit and integration tests to ensure the robustness and reliability of the applications.

My responsibilities:

- Development of Scala applications consuming/producing Kafka messages for parcel delivery

- Data integration via Spark, Scala, mainly with MySQL, PostgreSQL, and Cassandra

- Processing of parcel delivery events, with database integration via Kafka and Spark (60M events)

- Unit testing

- Implementation of observability

- Application deployment via Jenkins in production

Technical context: Scala, Kafka, Spark, MySQL, PostgreSQL, Cassandra, Jenkins, Kubernetes, ElasticSearch, Prometheus, Grafana