Jedify has launched Data Agents, letting business teams set up AI analysts in plain language and choose when they run. The notable shift is from asking a chatbot for an answer to having an analyst watch for a defined change and investigate it.
Watch Desk analysis
What happened
The product runs on Jedify’s context graph, which combines structured business data and documents with information about metrics, relationships and record-level permissions. Users describe the work they want done, then choose whether an agent runs on request, on a schedule or when a condition they define is met.
An agent can investigate a change such as a market starting to underperform. Jedify says data access follows the permissions of the person asking: information that person cannot see is excluded from the results. The agents can also be called through the Model Context Protocol from Claude, ChatGPT, Slack or another compatible platform. Data Agents is available now through Jedify’s Agents Hub, with a free starting option, SiliconANGLE reports. The article also describes RubyPlay using an agent to send a scheduled daily report.
Why it matters
The practical appeal is that teams may be able to hand off recurring analysis rather than repeatedly requesting the same dashboard or report. Scheduling and condition-based runs make these agents more like ongoing analysts than a chat window waiting for its next prompt.
Permissions are central to that promise. If the system respects the questioner’s existing access, it may help teams automate analysis without giving every agent a backstage pass to company data. That is Jedify’s account of how the product works; real-world reliability and the quality of its investigations remain to be demonstrated.
Our read
This is a specific, useful attempt to turn business context into repeatable analytical work. The strongest detail is not the “AI analyst” label, which has had a busy year, but the combination of scheduled or triggered runs and access tied to the user. We would judge it by whether the findings are accurate, useful and easy to check, not by how confidently an agent narrates a chart.
What to watch
- Whether users can verify an agent’s sources and reasoning behind an investigation.
- How condition-based runs behave when business data is incomplete or changes unexpectedly.
- Whether Jedify publishes results beyond the customer example in the launch report.
Discussion spark: For recurring business analysis, would you trust an agent to investigate and report automatically, or should a person have to start each run?
Sources and evidence
- Jedify formally launches Data Agents for building AI analysts in plain language – SiliconANGLE (28 October 2026, 13:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.