Jam is building a full-stack go-to-market agents team. Its agents take material from one platform and deliver it to another, for a wide range of users and use cases. After first building integrations in-house against APIs directly, Jam moved to Composio, which removed much of the prompt engineering each tool call needed and saved the team 40% of its engineering time on integrations.
Go-to-market agents work between platforms
Jam's agents send material across many different channels: they take content from one platform and deliver it to another. One example is its social monitoring workflow, which pulls posts from Reddit and X and sends them to Slack at whatever interval the user chooses.
Jam has a wide range of users and a wide range of go-to-market agents on the platform, so every one of those workflows depends on integrations with the platforms where the material lives and where it needs to go.
Direct API integrations meant prompt engineering for every tool call
Before Jam, the team had built integrations in-house for side projects, calling each platform's API directly. Jia found they had to spend a lot of time on prompt engineering for specific tool calls in very specific use cases.
That approach does not hold up on a platform where any user should be able to describe what they want and get a result in a single prompt.
Integrations aligned to the use case
For someone to prompt Jam and get results in one prompt, each use case needs direct alignment with the integrations behind it. That is what Jam found in Composio, and it eliminated much of the tool-specific prompt engineering the team used to do.
Composio saved Jam 40% of its engineering time on integrations, which Jia describes as quite a lot for the platform.
