The People Behind Starflow & QoD

Three open source engineers who believe data platforms should be declared, not coded. Together they build Starflow, the declarative pipeline platform, and QoD, the DuckDB serving gateway. Each owns a pillar: the vision, the SQL, and the orchestration.
3
Engineers behind Starflow & QoD
60+
Years of combined experience
100%
Open source, Apache-2.0
Hayssam Saleh

Founder & CTO

Hayssam Saleh

Creator of Starflow (formerly Starlake) and QoD, and founder of Starlake.ai. After two decades architecting large-scale data platforms for European enterprises, Hayssam distilled it into one conviction: pipelines should be declared, not coded. He now leads QoD into the agentic space, making AI agents first-class clients of the lakehouse rather than credential holders: identity-aware sessions, pinned snapshots, cost budgets, an auditable ledger, and writes that stay reviewable before they reach production.

  • Declarative Data Engineering
  • Agent-Native Data Platforms
  • DuckDB & Arrow Flight SQL
Andreas Reichel

SQL Architecture

Andreas Reichel

The engineer behind the Manticore Projects SQL ecosystem and everything SQL inside Starflow. Andreas maintains JSQLParser, the de facto standard Java SQL parser relied on by thousands of production systems, and authored JSQLTranspiler and JSQLFormatter on top of it. His work operates where few engineers go: parsing every major SQL dialect into a full abstract syntax tree, rewriting it, and re-emitting it for another engine. It is the reason a transformation written once in Starflow runs unchanged on DuckDB, BigQuery, Snowflake, Redshift, and Databricks, and what lets QoD serve warehouse-dialect SQL straight from DuckDB.

  • SQL Parsing & ASTs
  • Cross-Dialect Transpilation
  • JSQLParser / JSQLTranspiler
Stéphane Manciot

Orchestration & Distributed Systems

Stéphane Manciot

Founder of Softnetwork and architect of Starflow's orchestration layer. A veteran of distributed systems and reactive architectures, Stéphane turns declarative pipeline definitions into production-grade scheduled DAGs, inferring dependencies from lineage and generating native Airflow and Dagster workflows. His background in reactive, event-driven architectures is what keeps Starflow orchestration and QoD's self-healing node pools resilient, observable, and boring in the best possible way.

  • Workflow Orchestration
  • Airflow & Dagster
  • Reactive & Event-Driven Systems

Build With Us

Whether you want to deploy Starflow or QoD, contribute to the open source projects, or just talk data engineering, we would love to hear from you.

Get in Touch
Starlake.ai | Open Source Data Infrastructure for the Agent Era