Tabsdata - IT Press Tour #62 June 2025

Transcript

1. Pub/Sub for Tables: A Better Way to Ingest and Prep Data Arvind Prabhakar | Co-founder & CEO
2. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Founders Data integration leadership Alejandro Abdelnur Co-founder & CTO • 25+ Years of experience in enterprise software platforms • 10+ Years of experience in Data Integration and Analytics • First engineer and distinguished engineer at StreamSets • Early employee and engineering leader at Cloudera • Member of ASF and contributor to many Big Data projects Arvind Prabhakar Co-founder & CEO • 25+ Years of experience in enterprise software platforms • 15+ years of experience in Data Integration and Analytics • Founder, CPO/CTO of Streamsets from pre-revenue to ~ $100M • Early employee and engineering leader at Cloudera • Member of ASF and contributor to many Big Data projects
3. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Company background Journey so far… May 2024 Incorporated July 2024 Seed Funded February 2025 Patents Filed Out of Stealth July 2025 1.0 Release Public Beta June 2025 We are here
4. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Mission To make Pub/Sub the standard for data propagation – enabling enterprises to move faster, innovate with confidence, and turn every dataset into a trusted asset. Vision A future where data integration no longer exists – just trusted datasets, instantly accessible across the enterprise, ready for AI, analytics, and action.
5. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Data Preparation Ingest Why data pipelines fail to deliver on business outcomes Data pipelines are optimized for speed and volume, not for quality or trust… Databases Message Queues SaaS Applications Logs REST API Semistructured Unstructured Data Sources Data Normalization Cleansing Validation Aggregation Transformation Secure Data Storage Central Catalog Data Repository Data Consumers ML / AI Workloads Real-time Applications Business Intelligence Analytics & Data Science - New sources - Evolving semantics - Unclear ownership - Data bloat - Costly reprocessing and validations, false alarms, late reactions - Bolt-on tooling for quality, governance and metadata management - Changing requirements - Need for self-service - Perpetual phase lag
6. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. From pipeline-centric to product-centric data Data ownership, quality, and governance – built-in, not bolted on Sales Operations Data Engineer Sales Department Finance Operations Data Engineer Finance Department MySQL Technical Support Data Engineer Tech Support Department Analytics Data Engineer Analytics Department Subscriber Subscribes to Tables SF_OPPS SF_ACCOUNT Publisher Publishes Tables FI_LICENSE SU_LOGS SU_INCIDENT Customer Success Data Engineer CS Department CUSTOMER HEALTH CH_ENGAGEMENT CH_CONSUMPTION Transformer Derives new Tables
7. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Introducing Tabsdata: Pub/Sub for Tables A simple, scalable model to replace brittle and ineffective data ingest and prep… Secure Data Storage Central Catalog Data Repository Data Consumers ML / AI Workloads Real-time Applications Business Intelligence Analytics & Data Science - Reclaim agility - Clear ownership - Aligned with data strategy - No data bloat or reprocessing, false alarms etc - No need for bolt-on tooling for quality, governance etc - Built in provenance for complete traceability - Self-service access to data - Reduced time to value - Line of sight to data owners Sales Data Owners Technical Support Customer Success Finance Marketing Publish Data Contracts Data Products
8. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Systems that functionally overlap with Tabsdata You could achieve Pub/Sub for Tables using the following systems with some effort… databases workflow orchestrators message-brokers or event stores datalakes or data cloud
9. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Can you build Pub/Sub for Tables using a database? Yes but… databases • You will need to provide native connectivity to source and destination systems • You will need to build versioning of tables • You will need to build provenance functionality • And a bit more…
10. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Can you build Pub/Sub for Tables using a message broker? Yes but… • You will need to create table definitions using application level protocols • You will need to build versioning of tables • You will need to build provenance functionality • And a bit more… message-broker or event store
11. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Can you build Pub/Sub for Tables using an orchestrator? Yes but… • You will need to build data management for stage execution / execution state • You will need to build versioning of tables • You will need to build provenance functionality • And a bit more… workflow orchestrator
12. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Can you build Pub/Sub for Tables using a data platform? Yes but… • You will need provide datalake access to all data producers • Are you truly shifting left for quality and governance, or moving everything to the right? • Costs? datalakes or data cloud
13. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Tabsdata use-cases Simplifying data management and reducing operational costs • CDC for any data source • Automate data engineering • Simplify data integration • Implement data quality controls • Implement data contracts • Build data products • Improve governance
14. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Distribution • Open Core, with closed source enterprise extension • Free developer license • Targeted towards Python data engineers, available on PyPi • Self-managed on any infrastructure Pricing • Core based Go to market plan
15. Prepared for IT Press Tour | For Journalistic Use with Attribution © 2025 Tabsdata, inc. All rights reserved. Thank you!