How Qlik built grounded, enterprise-scale AI with Amazon Bedrock

Qlik built Qlik Answers, a grounded AI assistant for enterprise question-answering, on Amazon Bedrock, and the tool reached general availability in February 2026. Qlik says its Discovery Agent, an AI-driven anomaly and outlier detection agent, has surfaced more than 100,000 discoveries for customers since launch, and that the majority of Qlik Cloud accounts with agentic tools turned on are actively using them.

Qlik identified three problems to solve at once: orchestrating specialized reasoning without slowing users down, meeting data sovereignty requirements across Regions, and forecasting model capacity months ahead of demand. Qlik serves customers across Europe, Asia Pacific, and the Americas; it forecasts token consumption and model availability 3-6 months before major launches and validates projections against actual usage after rollouts.

The architecture uses distinct layers: an entry point inside Qlik Cloud; a lightweight routing layer; an answer layer that selects a fast or more deliberate path; a shared specialist-agent swarm runtime; a conversational analytics layer for structured data; a retrieval layer on Amazon OpenSearch Service; and a model access layer reached through Qlik's own LLM gateway connecting to Amazon Bedrock for chat, streaming, embeddings, and reranking. Bedrock Guardrails apply content filtering for prompt injection, PII, secrets, and denied topics, plus a grounding-validation check on generated answers. When a customer's model is not yet available in-Region through Bedrock, Qlik hosts it on Amazon SageMaker AI and later moves the workload back to Bedrock.

Qlik cites customer results reported separately: Bystronic deployed an AI chatbot in 15 minutes; Lintech International indexed more than 17,000 technical documents in Qlik Answers, cutting manual research time and speeding response times by 75 percent, returning as much as 7 hours a week to business managers; and TouchPoint Support Services, serving 650 healthcare sites and aiming to double in size, gives compliance-aligned guidance to 15,000 staff.

Qlik uses Amazon Bedrock cross-Region inference to serve 11 Regions while keeping data resident where compliance requires it. Qlik is evaluating Amazon Bedrock AgentCore for select workloads and is building a systematic framework for evaluating model performance.