Discover the Future of Salesforce Development
summary
Salesforce claims a new Agentforce studio and agentic layer that glue natural‑language roles, Data Cloud grounding, RAG retrievers and actions into deployable agents; the Atlas reasoning engine uses a react‑prompt system and learns over time. The developer toolchain adds Dev Assistant backed by homegrown LLMs codegen and xgen, plus prompt/model/retriever builders, Apex and Lightning Web Components support, sandboxes and CI/CD. Reported results: 75% usable test cases from generation and 276 billion Einstein predictions per quarter across the platform.
AI-written summary · about
chapters (8)
- 0:00Salesforce development intro
- 3:35Agentforce and AI agents
- 7:04Low-code to full-code tools
- 9:54AI-assisted Dev Assistant
- 14:09Why data matters for AI agents
- 16:56Agentforce components and workings
- 19:47Development future in human language
- 20:51Learning resources
related talks
07mSalesforce Platform & AI Agentforce as a CloudNative App
22mScaling AI for Developers: Lessons from the Inside
15mCode, Clouds, and AI: A Quick Look into Salesforce Developer Insights
01h 01mEnterprise Architecture 4 0 The AI Driven Future Preview
56mThe Age of the Marketing Maker is Here! Keynote (Replay)
32mFireside Chat with Richard Socher, Chief Scientist, Salesforce
33mTesting at the Speed of Agents: Salesforce's Evolving Approach to Agentic Testing
47mMarketing Keynote