Thomas B.

https://trainsphere.fr/ Built TrainSphere to help coaches and sports businesses acquire clients, showcase programs, and track performance through actionable analytics. End-to-end product built solo (with AI-assisted development), from discovery to production. • Conducted continuous user discovery with coaches to understand pain points (visibility, lead conversion, trust, admin time) and translate them into product requirements. • Designed and shipped the core experience + admin back-office: coach verification, moderation, secure reviews, referrals, gamification, and notifications — focused on trust and quality. • Delivered coach dashboards and built-in analytics (clickstream, timestamps, UTM, geo events) to help coaches understand what drives bookings and engagement, guiding product iterations. • Took full ownership of UX/UI and platform delivery: roadmap, prioritization, copy, support, and iteration loops based on feedback. • Technical implementation (high-level): Next.js/React/TypeScript, PostgreSQL + PostGIS (geo search + interactive map), deployed on AWS (containerized, production-grade). • Production mindset: CI/CD + IaC, monitoring/alerting, security controls, GDPR-ready cookie consent.

Long-term engagement modernizing a cloud data pipeline for solar park assets — now the single source of truth powering KPIs and near real-time monitoring. Weekly requirement refinement with an embedded Solar Engineer + close collaboration with PO and cross-teams. • Stabilized and improved the legacy AWS pipeline (Step Functions, Lambda, CDK/CloudFormation, MySQL, S3/JSON) while preparing the transition. • Led a dedicated HMI/web ingestion workstream: main interface for specs (data formats, validation rules, test strategy, acceptance criteria) and delivery coordination. • Owned the Airflow migration, rebuilding the pipeline as production-grade DAGs (scheduling, dependencies, retries, idempotency) with monitoring and run readiness. • Delivered the target platform: Airflow + Terraform + Python on EC2, S3 data lake + Delta Lake tables for scalable analytics. • Produced and maintained documentation (legacy + new), improving onboarding and long-term operability. • Acted as a cross-functional point of contact on data pipeline impacts, aligning constraints, timelines, and usability for consumers.

Versatile end-to-end role at the intersection of business needs (Marketing/Client Ops) and data engineering. Daily collaboration (often in pairing) with Marketing to clarify requirements, validate data definitions, and ensure deliveries remained aligned with evolving use cases. • Owned data quality and structure from acquisition/ingestion to external client delivery, ensuring datasets stayed usable, consistent, and business-ready. • Second to the CTO on the migration and full redesign of the company information system (IS). • Led the transition from locally / manually processed CSVs to a Serverless Data Factory on AWS. • Co-designed and implemented the processing chain: AWS Lambda (Python), DynamoDB, AWS Glue, SQS, SNS, Athena, S3, Elasticsearch. • Built operational reliability into the pipeline (monitoring, controls, repeatability): • Mean execution velocity: x7 • Mean defect rate: < 1% • Repeatability: 100% • Responsible for the design and implementation of GDPR compliance across the whole chain (data handling, access, traceability), in close alignment with business constraints.