Client engagement workspace vs AI receptionist

We get called an AI receptionist about once a week, usually by someone being helpful. I understand why. It is the nearest available category, and a new category is a tax you pay in every conversation until the market invents one for you.

But the two things solve different problems, and the difference is the whole reason the company exists. An AI receptionist answers the moment. A client engagement workspace manages what happens before and after it.

What does an AI receptionist usually do?

It answers, captures a message and sends it somewhere. The good ones qualify, book and transfer. That is genuinely valuable, and for a firm whose only problem is a ringing phone it may be the entire answer.

What it leaves behind is a handoff. Somebody at the firm now has a message, and the message is where the tool's job ends and the firm's job restarts. If the person who receives it is busy, in court, or simply not sure whether it is theirs, the message waits. The receptionist, human or otherwise, has done everything it was asked to do.

Where the losses actually happen

Most legal intake losses are not dramatic. They are ordinary. A message taken without context. A form sitting in an inbox. A callback nobody owns. A client explaining the same matter for the third time to the third person.

None of these look like failures while they are happening. That is what makes them expensive: there is no moment where somebody decides to lose the work. It leaks out through the joins between systems that each did their part correctly.

I spent years building the same joins for sales teams, where the problem is identical and the tooling is thirty years more mature. Legal has better people and worse plumbing, and the plumbing is the cheaper thing to fix.

What does a client engagement workspace do?

• Turns every channel into one enquiry record.
• Shows source, owner, urgency, status and next action.
• Recognises returning callers by the record the firm already has.
• Lets fee-earners ask what was promised and where the answer came from.
• Keeps the meeting note, the call and the follow-up on the same matter.
• Reports captured, recovered, booked and slipping work from the same records.

In other words, it manages the work that happens because somebody contacted the firm. The phone is only one entrance, and increasingly it is not even the busiest one.

Why the distinction is not just positioning

The test I would apply is this: can somebody at the firm, three weeks later, find out what was promised to a caller and by whom, without asking a colleague? If the answer is no, the firm has bought answering, not engagement, whatever the product is called.

That question also explains why the two categories price and behave so differently. Answering is priced per call or per minute, because the unit of value is the moment. A workspace is priced per firm, because the unit of value is the record, and a record gets more useful the longer it runs rather than more expensive.

Why this matters for legal AI generally

The legal AI companies firms trust do not feel like toys because they expose the control surface: sources, permissions, review, audit and workflow. Harvey and Legora built that discipline for documents, which is the right place to start, because documents are where the billable hours visibly sit.

We are building the same discipline for the part that happens before a matter is tidy enough to have documents at all. Somebody rings. Somebody fills in a form. Somebody sits in a consultation and is promised a call back. None of that is a document, all of it is the firm's relationship with the person paying, and almost none of it is currently written down anywhere a colleague can check.

Where I think the category lands

My honest expectation is that AI answering becomes a feature rather than a product. It is already close. The voice is solved, the cost is falling, and every case management vendor will ship a version of it because their customers will ask why they have to buy it separately.

What does not become a feature is the record. A record has to be built once, kept correct, permissioned properly and trusted by people who did not create it. That is slow to build and slow to copy, and it is the only part of this I would want to own in five years.

A receptionist captures the message. A workspace keeps the promise attached to it. That is why we talk about client engagement rather than reception: the job is not to sound like a person, it is to make sure the person at the firm has the record, the context and the next action in time to act on them.

Frequently asked questions

Do we have to choose between the two? Not immediately. Plenty of firms run an answering service and would keep it during a pilot. The question to ask after a month is where the messages go once they leave the service, and whether anybody can tell you.

Is this just a CRM for law firms? No, and we are careful about the word. A CRM records what your firm did. This records the whole conversation with a client across calls, messages and meetings, and it is designed so a fee-earner can ask it questions rather than maintain it.

Does it replace our case management system? No. The practice management system is deliberately the firm's system of record and a buyer should be suspicious of anyone claiming otherwise. WasilX sits in front of it and hands over once a matter exists.

What is the smallest useful version? One line, one office, and a weekly look at four numbers: captured, recovered, booked and slipping. If those do not move in six weeks, stop. That is a real answer and it is the one I would want if I were buying.