How Qlik built grounded, enterprise-scale AI with Amazon Bedrock
aws ·
Qlik developed Qlik Answers to help employees quickly retrieve trusted, sourced insights from vast enterprise data using Amazon Bedrock. The solution addresses common AI challenges by providing grounded answers that cite specific sources, ensuring accuracy and compliance in regulated environments. To manage complexity, Qlik utilized a modular architecture with distinct layers for routing, reasoning, and specialist agents, rather than relying on a single monolithic model. This design allows the system to handle diverse queries efficiently while maintaining data sovereignty across multiple global regions through Bedrock's cross-Region inference capabilities. Since its general availability in February 2026, the platform has facilitated over 100,000 discoveries via its Discovery Agent and demonstrated significant time savings for customers like Lintech International. Key architectural decisions included separating fast routing from deep reasoning and implementing strict grounding validation checks at the model access layer. By forecasting capacity months in advance and using feature flags for gradual rollout, Qlik ensured stable scaling without performance degradation. The case study highlights how combining flexible model selection with robust safety guardrails enables secure, enterprise-grade AI deployment.
aws