AI Project Assistant ARI: Live Context, Human Approval
A good project assistant knows the project, can do real work inside it, and stops where a person needs to decide. This is what ARI does, and what it deliberately does not.
ARI is CoMng.AI's in-project AI assistant. It has the live context of the project (tasks, milestones and their conflicts, budget, team, decisions), drafts change requests and notes, converts customer insights, and runs plan checks, but it never applies a change without human approval.
- Full project context: tasks, milestones and their dependencies, detected schedule conflicts, budget, team, activities, and every Project Settings list.
- Does real work: drafts change requests and notes, converts insights and emails, posts task comments, runs plan coverage checks, calculates priorities.
- Stops at the gate: ARI never applies a change itself; approval is a human step.
- Honest limits: it cannot act outside the platform: no email, calendar or monitoring.
What ARI Is
ARI is a conversational AI teammate available from the header on any project page. Unlike a general chatbot you paste context into, ARI already has the project: its tasks, milestones, budget, team, activities and notes. Ask it a question and the answer is grounded in your data.
What ARI Knows
- Tasks and subtasks, including blocked status and reasons, priorities and tags.
- Milestones and confirmed milestone dependencies, plus any detected planning conflicts (a task scheduled after a milestone it gates). Scheduling questions and change requests account for these without being asked.
- Every Project Settings list: goals, success metrics, risks, stakeholders, scope, requirements, assumptions and issues, so it can answer "what did we decide about X?" from the recorded decisions.
- Budget lines and payments, team members and activities.
What ARI Can Do
| Capability | Detail |
|---|---|
| Draft change requests | Modify tasks (incl. subtasks, blocking and blocked-reason, tags), milestones, budget lines and payments, risks, team members, scope, notes, goals, activities, dependencies and the other Project Settings lists, with before/after values |
| Write knowledge | Draft knowledge-base notes |
| Convert inbound items | Turn a customer insight or inbox email into a task, note, budget item or AI-drafted change request |
| Collaborate | Post task discussion comments |
| Check the plan | Run Verify Plan Coverage (check-only by default) and calculate priorities in bulk |
| Work on the whiteboard | Generate ideas, write documents from the board, and create tasks |
Creating Tasks With ARI (Through a Change Request)
Ask ARI for the work you need and it drafts a change request that adds the tasks, along with any related changes, showing before and after values in the sandbox. ARI does not create the tasks directly; you review the request and approve it, which keeps AI-generated work under the same controls as everything else. For a bulk generation from project data or from a written brief, the Tasks page offers the AI Task Builder and AI Brief, each with the same apply-or-review choice. See Task Capture.
Where ARI Stops, By Design
ARI never applies anything itself. Drafted changes land as change requests that a person reviews in a sandbox and approves. Bulk or destructive edits are additionally flagged by a deterministic check, not by the model's judgment. And ARI is explicit about its limits: it cannot send emails, manage a calendar or monitor systems outside the platform. If it cannot do something, it says what is blocking it rather than implying it did. See AI Governance and Safety Gates.
Four Workflows Worth Trying First
Ask why
"Why is the Data Mapping milestone at risk?" ARI points to the late task and the days involved.
Ask what we decided
"What did we decide about the cutover date?" ARI answers from the recorded decisions.
Draft the fix
"Draft a change request that moves the milestone." You get a sandboxed request to review.
Clear the inbox
"Convert this customer email into a task." ARI uses the same logic as the manual Convert button.
- Copy project data into a chatbot
- Answers without the plan's context
- Edits made by hand afterwards
- The assistant already has the project
- Answers grounded in tasks, decisions and conflicts
- Drafted changes reviewed and approved in-app
About CoMng.AI
CoMng.AI is an AI-native project management platform that generates a project's plan, tasks, risks and budget from a description or document, monitors project health in real time, and gives both human teams and AI coding agents live access to the same project data, with a person approving every AI-generated change.
See AI propose changes while a person stays in control
Open a project, ask ARI to draft a change request, and review the sandbox preview before anything is applied.
Frequently Asked Questions
What is ARI in CoMng.AI?
ARI is the in-project AI assistant. It has the project's live context and can answer questions, draft change requests and notes, convert insights and emails, and run plan checks, but it never applies changes without approval.
Can ARI edit my tasks directly?
ARI drafts changes as change requests that a person approves and applies. A few actions are done directly and are logged, such as posting task comments or calculating priorities in bulk.
What can ARI not do?
It cannot act outside the platform: it cannot send email, manage calendars or monitor external systems. It tells you when something is outside its tools.
Does ARI know about decisions we made?
Yes. It sees every Project Settings list, including the Assumptions tab where decisions are recorded, as well as pending change requests and meeting notes.
Is ARI available on mobile?
The mobile app includes Ask AI. Deep AI tooling such as change requests is desktop-only today.
Can ARI create tasks for me?
ARI can draft a change request that adds tasks, with before and after values and a sandbox preview. A person approves it before anything is applied. For bulk generation use the AI Task Builder or AI Brief on the Tasks page.
Related reading
See the full guide: AI Agents, ARI & MCP Guide