Search with citations
One query across your document stores, HRIS, ERP, and collaboration tools. Every result carries its source and the user's existing permissions apply.
Platform
Assist is the conversational interface across Meridian and your connected systems. Search with citations, generate documents from governed data, and start agents from the same place, with your permissions applied.
What drove the Q3 opex variance for Field operations, and which lines are over plan?
Planning Agent · read finance.gl_actuals and finance.plan as d.okafor · Snowflake, 4 min old · 1.8 s
Field operations is $268,412.55 over plan for Q3 2026 (+6.5%). Three lines account for 91% of the variance.
| Line | Variance | |||
|---|---|---|---|---|
| Contract labor | 620,000.00 | 764,300.00 | +144,300.00 | 3 unbudgeted contractors, Jul to Sep |
| Overtime | 410,000.00 | 476,918.20 | +66,918.20 | Peak-season shift coverage |
| Fleet maintenance | 285,000.00 | 318,206.35 | +33,206.35 | Q2 service deferred into Q3 |
gl_actuals · 2026-09-08plan_v3 · FY26
Acting as D. Okafor · Northwind Logistics · your permissions apply
Most work starts with a question. Where is the latest version of the expense policy. What did we pay this vendor last year. Draft a job description for a senior accountant in Dublin at the L4 band. Assist answers from the systems you have connected, cites its sources, and, when the answer is an action, starts the right agent.
Assist respects permissions at every step. Search results include only documents and records the user can already open. Generated documents draw only on data the user can already query. Agents started from Assist run under the user's identity with the approval steps they always have.
It works in the browser, in Slack, and in Microsoft Teams, so employees, managers, and finance teams meet the platform where they already work. For HR service delivery, the Help Desk Agent runs inside Assist; for everything else, Assist is how people find things and get them started.
Capabilities
5 capabilities, each enforced below the agent so they apply to Meridian, partner, and Studio-built agents alike.
One query across your document stores, HRIS, ERP, and collaboration tools. Every result carries its source and the user's existing permissions apply.
Job descriptions, policy summaries, variance narratives, and board-ready tables generated from data the user can already see.
Start the Audit Agent on a PBC list or the Scheduling Agent on an open shift from the same conversation. Approvals still apply.
Full functionality in the tools people already use, plus the browser workspace for longer tasks.
Assist never holds its own access. It acts as the user, through the Gateway and the Data Fabric, so answers cannot leak across roles.
Specification
Standards, protocols, and limits as of September 2026. Your account team can confirm coverage for a specific stack.
| Item | Specification |
|---|---|
| Surfaces | Browser workspace, Slack, and Microsoft Teams |
| Search | One query across document stores, HRIS, ERP, and collaboration tools; every result cites its source |
| Generation | Job descriptions, policy summaries, variance narratives, and tables from data the user can already see |
| Agent runs | Start any registered agent from the conversation under the user's identity; approval steps still apply |
| Permissions | Acts as the user through the Gateway and Data Fabric; holds no access of its own |
| Models | Workspace-selected frontier models; provider and region can be pinned; domain reasoning routed to Meridian's tuned models |
| Logs | Conversation logs retained under your policy and visible to the user |
| Plan availability | Starter, Growth, and Enterprise |
How it fits
One identity model, one policy layer, one log. The other components and the trust model that binds them.
Questions
It is a conversational interface, but it is not an open-ended assistant. It answers only from your connected systems with citations, generates only from data the user can see, and starts governed agents rather than improvising actions. When it cannot answer from your data it says so.
The models your workspace administrator selects. Meridian supports frontier models from leading providers and routes HR and finance domain reasoning to Meridian's own tuned models. Your data is not used to train any of them.
Yes. The Help Desk Agent runs inside Assist. Employees ask about pay, leave, benefits, and policy in Slack, Teams, or the browser, and the agent answers from policy and their own record.
No. Assist acts as the user through the Gateway and Data Fabric. Search, generation, and agent runs all carry the user's identity and permissions.
Get started
Start with one workflow, one approver, and one number to move. Most design partners were live in five weeks.