AWS re:Invent 2025 - All Power, No Pain: Scale Agentic AI Apps in an Open Lakehouse (AIM247)

Summary of AWS re:Invent 2025 Presentation: "All Power, No Pain: Scale Agentic AI Apps in an Open Lakehouse"

Introduction to Click and AWS Partnership

• Click is a global software company that provides an end-to-end data platform for data integration, transformation, and analytics.
• Click has a long-standing strategic partnership with AWS, leveraging AWS services to power their data platform capabilities.
• The goal of the Click-AWS partnership is to help customers deliver successful data projects on AWS as quickly and easily as possible.

The Current AI Landscape

• AI initiatives are becoming more widespread, with nearly 70% of organizations having a formalized AI strategy.
• However, organizations are struggling to predict and quantify the ROI of AI projects, slowing down adoption.
• A key barrier to successful AI projects is the lack of a strong, trustworthy data foundation.

Building an Open Data Architecture

• Click believes that an open data architecture, centered around an open lakehouse approach, is the key to delivering successful data and AI initiatives.
• The open data architecture should support a wide range of data sources, from simple APIs to complex streaming and transactional systems.
• The data platform should provide capabilities for efficient data movement, transformation, and analysis to enable high-quality data for various consumers, including AI applications.

The Benefits of an Open Lakehouse Approach

• Cost Efficiency: Lakehouse technologies, specifically those based on the Iceberg format, have proven to deliver flexible data storage at a lower cost compared to traditional data warehouse approaches.
• Interoperability: Iceberg has emerged as a widely adopted lakehouse format, enabling interoperability across the data ecosystem, including AWS services and various data consumption engines.

The Click Open Lakehouse Solution

• Click's Open Lakehouse solution, powered by Apache Iceberg and running on AWS, provides three key capabilities:
  1. High-Throughput Ingestion: Ingests data from a wide range of sources, including real-time streaming data, into Iceberg tables.
  2. Adaptive Optimization: Continuously optimizes Iceberg tables for low cost and high performance, handling tasks like file compaction and snapshot management.
  3. Warehouse Mirroring: Presents the optimized Iceberg tables to a variety of consumption engines, enabling interoperability and open access to the data.

Performance and Cost Benchmarks

• Click conducted a benchmark comparing real-time streaming data ingestion into their Open Lakehouse solution versus a traditional data warehouse.
• The results showed that the Open Lakehouse solution delivered:
  • 5x faster data freshness (1-3 minutes vs. 5-15 minutes)
  • 75-90% lower cost compared to the data warehouse

Demonstration and Real-World Example

• Click demonstrated a real-time streaming data ingestion use case, where sensor data from IoT-enabled bikes was ingested into Iceberg tables using the Click Open Lakehouse solution.
• The demo showcased the end-to-end pipeline, from Kinesis data ingestion to Iceberg table optimization and querying using tools like AWS Athena.
• The data quality was also highlighted, with the Click platform providing data quality metrics and trust scores to ensure the ingested data is accurate and reliable.

Key Takeaways

• An open data architecture, centered around an open lakehouse approach, is crucial for delivering successful data and AI initiatives.
• The Click Open Lakehouse solution, powered by Apache Iceberg and running on AWS, provides a cost-effective and scalable platform for ingesting, optimizing, and accessing high-quality data.
• The solution's performance and cost benchmarks demonstrate significant improvements over traditional data warehouse approaches, making it an attractive option for organizations looking to build a strong data foundation for their AI and analytics initiatives.