About us and the role:

We build a fully open-source data-mesh platform that lets organisations, institutions, and states share data securely and keep control over it — for health systems and the public sector across Africa and Europe, including work for institutions such as the African Union. This is real data and platform engineering with continental impact, not generic website work.

We are growing our engineering team and hiring a Mid-Level Data Engineer / Full-Stack Developer, based at our Speedykom MEA office in Tunis. You will split your time roughly evenly between building and running data pipelines and analytical stores, and developing the full-stack application that exposes data products, pipelines, and dashboards to users. This is an on-site role in Tunis (not remote-first), and it involves regular travel across Africa: beyond writing code, you will run workshops and deliver training to partner and public-sector teams, so communication and critical thinking are central.

The technical environment:

On the data side you will work with ClickHouse as our analytical store, Apache Airflow and Apache Hop for orchestration and ETL, MinIO / S3 object storage and Parquet, DuckDB, data contracts, DataHub for cataloguing and lineage, and Apache Superset for BI. On the application side, a Python / Django REST backend and a Next.js / React / TypeScript frontend, with PostgreSQL and Redis, containerised with Docker and deployed on Kubernetes, secured with Keycloak, and shipped through GitHub Actions.

What you will do (about 50% data, 50% full-stack):

Design, build, and maintain data pipelines: ingest data into object storage (MinIO / S3, Parquet), transform it, and load it into ClickHouse for analytics.

• Orchestrate and schedule workloads with Apache Airflow and Apache Hop, and make them observable, reliable, and re-runnable.
• Model data products and data contracts, and help evolve the catalogue and lineage in DataHub.
• Build end-to-end features across the Django / Python (DRF) backend and the Next.js / React / TypeScript frontend — data catalogue screens, pipeline interfaces, and dashboards.
• Design and consume REST APIs, work with PostgreSQL, and integrate analytics and BI through Apache Superset.
• Containerise services with Docker, contribute to CI/CD pipelines (GitHub Actions), and write tests and clear documentation.
• Apply secure-coding basics and identity & access management (IAM / SSO with Keycloak; OAuth2 / OIDC; RBAC), secrets handling, and OWASP fundamentals in everyday work.
• Travel to partners across Africa and deliver workshops and training on the platform and its data products — this role is more than writing code.

• Your profile — required:
  • At least 3 years of professional experience after graduation. Internships, PFE / final-year projects, and working-student roles do not count towards this.
  • Excellent communication and strong critical thinking — decisive for this role. You will explain, question, present, and train, not only code.
  • Willingness and ability to travel within Africa for on-site work, workshops, and training.
  • Strong command of Python and SQL.
  • Experience with Django / Python or a comparable backend framework, and with building and consuming REST APIs.
  • Genuine data-engineering fundamentals: ETL / ELT, data modelling, working with analytical / columnar stores, and handling large data volumes.
  • Comfortable with a modern JavaScript framework (Next.js / React) and TypeScript.
  • Comfortable with PostgreSQL, Git / GitHub, and Linux, and a solid grasp of Docker and containerisation / virtualisation.
  • Proficiency with AI-assisted development and modern coding tools, and solid familiarity with software-engineering best practices.
  • A real security-first mindset: secure coding and awareness of common vulnerabilities (OWASP).
  • Genuine understanding of and enthusiasm for open-source software and how to work within open-source ecosystems.
  • Strong written and spoken communication in both English and French. You will work in an international team, train partners, and represent us to public-sector teams, so clear communication in both languages is essential.
  Strong pluses (not required):
  • Hands-on experience with ClickHouse or another OLAP / columnar database.
  • Experience with Apache Airflow, Apache Hop, or another orchestration tool.
  • Experience with MinIO / S3, Parquet, DuckDB, data contracts, DataHub, Kafka, or Apache Superset.
  • Exposure to Kubernetes and observability (Prometheus / Grafana / Loki).
  • Hands-on Keycloak / IAM (SSO, OAuth2 / OIDC, RBAC).
  • Experience running workshops or training, or in a client- / partner-facing role.
  How we work and what we expect:

  You read requirements carefully, ask precise questions, and implement features end-to-end without waiting for micro-management. You take ownership and can structure your own work. Above all, critical thinking and communication are decisive: you will not only ship code — you will travel to partners across Africa, run workshops, and deliver training. Fluency with AI and modern coding tools and familiarity with best practices matter, but your ability to reason clearly and communicate well matters most. We expect you to use AI tools to raise the quality and speed of your work — not to do less; you remain fully accountable for the result. We want people who deliver quickly and to a high standard, bring genuine enthusiasm, inspire the people around them, and show real commitment and team spirit.

  What we offer
  • A permanent contract (CDI) through Speedykom MEA, based at our Tunis office.
  • Real data and platform engineering on systems with continental reach — not maintenance of generic sites.
  • Exposure to partners and public-sector teams across Africa, with travel and training as part of the role.
  • A team that works seriously on open source, data sovereignty, and DevSecOps.
  • Room to grow across the full data and application stack, working with modern open-source tools.
  Selection process:

  A short set of screening questions (including your graduation year and employers since then), followed by a focused technical conversation and a small, time-boxed task — no long unpaid take-home. One-click applications without answers to the screening questions will not be processed.