LangChain built Content Studio, an open-source design agent in Slack. It creates on-brand videos, graphics, and decks from a single prompt. The agent runs on Managed Deep Agents and uses a sandboxed environment per thread. It has delivered over 3,000 assets across the company. Posts with its work earned 4.5M views on X. LangChain posts with its visuals saw 2x likes and 3.7x bookmarks. Three in four requests now come from outside marketing. The median request costs about 50 cents in model spend. The open-source repo and harness are available for forking. The team reduced their own code by about 80% after migrating to Managed Deep Agents.
What changed
Design requests were handled manually by an overstretched design team.
What it unlocks
Non-designers can now generate on-brand videos, graphics, and decks from a Slack prompt.
- 3,000+ assets delivered
- 80+ slide decks created
- 4.5M views on X
- 2x likes on LangChain posts
What you need to act on it
- Managed Deep Agents deployment
- Slack integration
- brand tokens, fonts, and skills configured
Sources