How to put an AI agent inside a CRM with a person approving every action
An AI agent fits a customer-facing team when it drafts the work and a person approves it, inside the CRM the team already uses. At Canam Consultants, that design now handles 80% of repetitive counselor tasks.
Intelligent systems2 min read
A practical way to put an AI agent into a customer-facing team is to let it prepare the work and have a person approve each action before it happens. The agent lives inside the CRM the team already uses, so the approval is one click in a familiar screen, and every action carries a record of who approved it.
We built an AI agent into Canam Consultants' CRM, with a counselor approving each task. The agent now does 80% of the repetitive counselor tasks. Counselor capacity doubled, and student satisfaction stands at 4.8 out of 5.
80%
of repetitive counselor tasks at Canam Consultants are done by an AI agent in the CRM, each approved by a person
Pick tasks that repeat and have a clear right answer
Start by listing what the team does many times a day: follow-up messages, document requests, status updates, reminders before a deadline, routine replies. Keep the tasks where a reviewer can tell at a glance whether the draft is right. Leave out anything that needs a judgment call about a person's situation; those stay with the counselor.
Put the agent where the work already happens
An agent with its own app becomes one more tab nobody opens. Build it into the CRM: the draft appears on the record it concerns, with the data it used, and the approve, edit and reject buttons sit next to it. The agent reads from the CRM and writes back to it, under the same permissions as the person approving.
Make approval fast and rejection useful
If approving takes longer than doing the task by hand, people stop using the agent. Show the draft, the reason it was proposed and the record it will change, on one screen. When a counselor edits or rejects a draft, store the change. Those corrections are the best data you have for improving the agent, and they show you which tasks it should not attempt yet.
Record every action for audit
Each action needs a trail: what the agent proposed, what data it read, who approved it, what changed and when. In a regulated business this is what an auditor will ask for. It also settles disputes with customers, because you can show exactly what was sent and who sent it.
Widen the scope only on evidence
Track the approval rate per task type. A task approved without edits almost every time is a sign that similar tasks can be added. A task that keeps getting rewritten goes back to the team. The scope grows from the record, not from a roadmap written before launch.
Questions
- What is a human-in-the-loop AI agent?
- An agent that prepares an action, such as a message or a record update, and waits for a person to approve, edit or reject it before anything happens.
- Should an AI agent have its own interface or live in the CRM?
- In the CRM. Drafts should appear on the record they concern, with approve, edit and reject next to them, so the team does not switch tools.
- What results did Canam Consultants see from its CRM agent?
- The agent does 80% of repetitive counselor tasks, each approved by a counselor. Counselor capacity doubled and student satisfaction is 4.8 out of 5.
More on intelligent systems
All writingNovember 2026
CERT-In's AI security audit guidelines: what an auditor will check
CERT-In's AI Security Auditing Guidelines, dated November 4, 2026, sort AI systems into four risk tiers and audit them across nine domains. For AI agents, auditors will test whether an agent can exceed its permissions or bypass approval. Here is what to prepare.
October 2026
How do you test an LLM before it goes live in a regulated business?
Test a language model the way an auditor will question it: on your own cases, against written pass criteria, with attacks on its permissions and data, and with a record of every result. RBI's Governor has asked banks to red-team and stress-test AI before deployment and periodically after; five test sets are where that starts.
October 2026
Multilingual voice AI for Indian call centers: a build checklist
A voice AI line in Indian languages succeeds when it resolves the calls that make up most of the volume, hands the rest to a person with context, and logs every call in the core systems. A build checklist, with results from a line we built that answers in 11 languages.
Tell us which system the AI has to run inside
If that system is not ready for AI yet, tell us that too. An engineer will reply and say plainly whether we are the right firm for it.