Clarita: Clariant’s journey to accelerate growth with GenAI and RAG(AIM305)
summary
Clariant builds Clarita, a RAG-powered assistant that indexes ~400 TB of corporate know‑how into a vector store and returns synthesized answers in about 30–40 seconds; the platform uses Amazon Bedrock and Amazon OpenSearch for model inference and retrieval, an orchestration layer on Lambda calling agents and embeddings, and automated evaluation to raise quality over time. Adoption metrics show ~40% of employees tried it and ~30% use it monthly, and use cases span R&D literature search to sales briefs.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 0:51Why customize models
- 2:43Customization approaches
- 5:03Understanding RAG
- 6:53Semantic search mechanics
- 8:34Embedding models deep dive
- 12:16Implementing RAG
- 22:15The Clarita platform
- 24:01Business applications
- 26:52Clariant's AI journey
- 34:32RAG in action at Clariant
- 36:28Data pipeline architecture
- 38:42Key learnings and mindset
- 44:18Strategic summary
- 45:18Conclusion and resources
related talks
27mTransforming Financial Services with Generative AI
21marchive – Clara Zang
51mTransforming Banking with Advanced AI: GenAI, RAG, and Multi-Agent Systems
40mRolling out Enterprise AI: Tools, Insights, & Team Empowerment
35mThe Hidden Jewels to Solve Confabulation Challenges in Generative AI’s RAG Promise
29mCustomizing genAI applications for your business
38mScaling Specialist Knowledge with AI
27mRio Tinto unlocks potential of generative AI