Keynote: Enterprise Value with Generative AI
Talkat Cassandra Summit 2023 · on Tuesday, 2023-12-12· at 09:30· Grand Ballroom BC (Concourse Exhibit Level)·
summary
Claims that LLM development explodes experiment dimensionality and produces massive generated metadata, so teams need a system of record tying datasets, model variants, prompts and outcomes; calls for orchestration and continuous monitoring to make experiments repeatable and control costly compute; highlights lack of a standard metric for hallucination and the need for flexible evaluation, documentation and governance; proposes a modular toolset—experiments/sweeps, artifacts, launch, prompts and monitoring—from Weights & Biases (free tier available) as a pragmatic way to capture and reproduce LLM work.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:06Enterprise value with generative AI
- 2:32Generative AI tooling challenges
- 3:13Volume of generated metadata
- 4:15Orchestration complexity
- 5:59Evaluation and governance
- 8:48Weights & Biases toolkit vision
- 9:51Q&A
related talks
10mKeynote: Enterprise Value with Generative AI
40mKeynote
40mPannel Discussion | From Hype To Value: Real World Applications Of Generative AI And LLMs
32mBeyond LLMs: Building an AI-Powered Decision System at Scale
12mKeynote: How LLMs Have Changed the Way Organizations Do AI
33mOpening Keynote: LLMs, between fantasy and reality - the poin...
15mKeynote: Into the Black Box: Observability in the Age of LLMs
29mGenerative-AI: Usecase-Specific Evaluation of LLM-powered Applications