Roland Saïkali
Senior Software Engineer · Python Backend & Data Systems
Python · FastAPI · PostgreSQL · Time-series · Distributed systems · Data pipelines
Senior software engineer, 25+ years building Python, data-intensive systems for large French industrial and public-sector platforms: PagesJaunes (Cassandra clusters serving 100,000+ sites), Airbus (simulation-data frameworks), EDF, Deutsche Telekom and others. Designs and ships full backend chains: REST APIs, time-series storage (PostgreSQL, TimescaleDB, InfluxDB), data pipelines and operator-facing UIs. Recent independent work applies the same stack to energy data and applied machine learning, including a NILM pipeline (TensorFlow / PyTorch) disaggregating appliance usage from smart-meter signals. Co-author of itinerennes.info, winner of the Rennes Open Data Grand Prix du Jury.
- Python and data engineering for client projects, alongside self-directed R&D in energy data and applied machine learning.
- Python / data engineering consulting for clients: backend services, REST APIs, data pipelines, ETL and time-series processing.
- Built a NILM (Non-Intrusive Load Monitoring) pipeline disaggregating appliance-level consumption from a single smart-meter signal, using TensorFlow / PyTorch with pandas and scikit-learn, served as an inference microservice.
- Runs a production-grade data platform end to end: ingestion, TimescaleDB / InfluxDB time-series storage, Grafana dashboards, automated deployment on Docker Compose.
- Published 15+ open-source Python projects (github.com/rsaikali): data ingestion bridges, sensor integrations and time-series ETL.
- Microsoft Teams chatbot platform. Azure DevOps deployment automation, monitoring, alerting and quality assurance.
- Python / data consulting for local clients, with self-directed R&D building the first generation of an energy-data and time-series platform.
- Python / data engineering consulting for local clients.
- Designed the first generation of an energy-data platform: smart-meter (Linky / TIC) ingestion, custom Python services, early time-series storage and dashboards on a distributed Raspberry Pi cluster.
- Built and open-sourced foundational Python tooling reused across later projects.
- Infrastructure-as-Code datacenter orchestration platform for Deutsche Telekom. Designed a declarative model where services, hardware and networking are described as code, compiled into reproducible virtual environments for prototyping, testing and production deployment.
- 15 years of software-engineering consulting for large French industrial and public-sector clients. Selected assignments:
- PagesJaunes / Solocal (2009-2015): operated Cassandra NoSQL clusters and Karaf application servers, built Python / Fabric automation tooling and Jenkins CI/CD, optimised the Pertimm search engine on a platform hosting 100,000+ websites.
- Airbus Aerocity (2004-2008): designed a Python framework and data-management API for wind-tunnel simulation data.
- Earlier assignments: EDF, IRSN, Airbus (JEDM, LTS), BioCoop, Generali, CNIEG. Software engineering across Python, Java, ETL (Talend) and Visual Basic stacks.