From Open-Source to SaaS with Head of Engineering at Onehouse
Varun Madan, Head of Engineering at Onehouse, walks through the journey of transforming the open-source Apache Hudi project into a high-performance, enterprise-ready SaaS platform. Covers data lakehouse architecture, building SaaS offerings, and includes a live demo of real-time data ingestion.
Special Guests
Varun Madan
Head of Engineering at Onehouse, transforming Apache Hudi into enterprise SaaS.
Timestamps
00:00Intro: Data Quality for Agentic AI01:15What is a Data Lakehouse & Why It Matters03:50Apache Hudi: From Uber to Open Source05:45Lake vs Warehouse vs Lakehouse08:00Key Lakehouse Features: Indexes, Metadata, Speed10:20From Open Source to SaaS: Building Onehouse13:00Solving Real-World Enterprise Data Challenges16:00Live Demo: Ingesting Kafka Data into Onehouse19:30Running AI Workloads with Ray21:30Final Thoughts & How to Get StartedAbout This Episode
- What is "From Open-Source to SaaS with Head of Engineering at Onehouse" about?
- Varun Madan, Head of Engineering at Onehouse, walks through the journey of transforming the open-source Apache Hudi project into a high-performance, enterprise-ready SaaS platform. Covers data lakehouse architecture, building SaaS offerings, and includes a live demo of real-time data ingestion.
- What topics does this episode cover?
- This episode is filed under Open Source, Data & Databases, SaaS Strategy on Agentic SaaS Talks.
- How long is this episode, and when was it published?
- "From Open-Source to SaaS with Head of Engineering at Onehouse" runs 23 min and was published on October 10, 2025.
- Who is featured in this episode?
- This episode features Varun Madan as a guest, in conversation with the Agentic SaaS Talks hosts.
- What are the key moments in this episode?
- Key moments include: What is a Data Lakehouse & Why It Matters; Apache Hudi: From Uber to Open Source; Lake vs Warehouse vs Lakehouse; Key Lakehouse Features: Indexes, Metadata, Speed; From Open Source to SaaS: Building Onehouse; Solving Real-World Enterprise Data Challenges; Live Demo: Ingesting Kafka Data into Onehouse; Running AI Workloads with Ray. The timestamps on this page jump to each one on YouTube.
Related Episodes
Episode 35: The Founder's Guide to Scaling Open Source AI Products to Production on AWS
Omnistrate co-founders Kamal Gupta and Alok Nikhil map the road from open source adoption to revenue: the four ingredients of monetization (project, enterprise features, distribution, go-to-market) and the three business models open to open source AI companies, product-led growth, AWS Marketplace co-sell, and enterprise sales. Alok closes with a live demo that turns vLLM into a managed inference service with a self-service customer portal, showing why distribution needs a control plane and why a Marketplace listing alone gets you discovered but not paid.
Episode 28: Commercializing Open Source: Why Most Founders Get It Wrong
Recorded at the AWS re:Invent SaaS Summit, four open-source veterans from Percona, Couchbase, and Omnistrate break down what actually works when commercializing open source, and the expensive mistakes founders make. The panel covers open core vs. foundation models, avoiding MongoDB-style license backlash, and partnering with hyperscalers without getting 'Amazon'd'.
Episode 25: Cloud-Native Databases on Kubernetes for AI Agents, with Sergey Pronin
Sergey Pronin, founder of Solanica and former Percona engineer, joins the panel to discuss why the SaaS-only data model is fracturing. The conversation covers 'Dark Matter Data' trapped inside enterprises, treating databases like cattle on Kubernetes, API-driven self-service provisioning, and why donating Open Everest to the CNCF matters for license-stable infrastructure.