Almost every company building voice AI is working on the same worry: will the person on the other end realise it is a machine, and how badly will that go.
Wayco has found the calls where nobody minds.
What they do
Wayco runs a 24 hour AI operator for law firms, starting with personal injury and the medical work around it. It handles taking on new clients, case referrals, and the coordination between legal teams, medical providers and the claimants themselves. Work that traditionally takes days of phone calls, the company says, comes down to minutes.
It is worth being concrete about what that coordination is, because the word hides the problem. A personal injury case cannot move unless three parties keep passing information between them. The law firm needs to know what treatment happened and when. The clinic needs approvals, appointments and record requests handled. The claimant needs appointments, updates and documents. None of them share a system. So it runs on the telephone: someone calls, someone is out, a message is left, a day goes by, somebody calls back.
The company raised a $500,000 seed, backed by Pioneer Fund. Its founder, Iqbol Temirkhojaev, is 19, from Tajikistan, and has been shipping venture-backed software since he was 13. The United Nations acquired one of his companies when he was 14. He held a software patent at 15.
Who else is in this field
Law firms have been sold software for decades. Clio and Filevine run the case itself. Lawmatics handles taking on new clients. EvenUp works specifically on personal injury paperwork. On answering the phone, Smith.ai, Ruby and Answering Legal have all sold human receptionists to firms for years. The general-purpose voice agent companies, Vapi, Retell AI and Bland, will sell the underlying technology to anyone.
Answering the firm's own phone is well covered. Doing the calling out, between the firm and the clinics, is not.
The point: these are the calls where nobody minds an AI
The reason voice AI has been hard is not really technical. It is that most of the calls people have tried to automate are calls where the human wanted a human: a complaint, a sale, a negotiation, anything where the other person is forming a view of you while you speak.
The calls Wayco handles are not like that. A records clerk at an orthopaedic clinic dealing with a request for a treatment summary is not looking for a relationship. A scheduler confirming a date wants the date confirmed. The claimant chasing an update wants the update.
Nobody on these calls is hoping to connect with someone. They are trying to finish a task, and an AI that says plainly what it is and then finishes the task is not a disappointment. It is a relief.
That one fact removes the hardest problem in the category. Wayco does not need to pass as human, does not need to manage a disclosure, and does not inherit the awkward ethics of pretending. It can be obviously a machine and still be entirely welcome, because it is there at seven in the morning, it never has a full voicemail box, and it will call back an eleventh time without sounding tired of you.
Coordination nobody can opt out of
The second thing that makes this a good place to stand is that the communication is compulsory.
Most software sells into things people could choose not to do. You could send the newsletter or not, run the campaign or not, hold the meeting or not. The coordination in one of these cases is not optional. The case does not progress without it. The records must be requested, the treatment must be documented, the claimant must be reachable, or there is no claim.
So the call volume exists whether or not anybody buys anything, and it will keep existing. A vendor here is not creating demand. It is standing in front of work that is already happening badly.
There is a knock-on effect worth noting. Because the operator sits between the parties rather than inside one of them, it ends up carrying the traffic of the referral relationships personal injury runs on. Whoever reliably moves the messages between a firm and the clinics it works with becomes, over time, hard to route around.
Why a law firm actually buys it
The pitch to a managing partner is not that the software is clever. It is competitive.
Personal injury is a crowded, heavily advertised, referral-driven trade where firms compete on how quickly they respond. The firm that picks up at nine on a Sunday signs the client. The firm that gets medical records back in four days settles sooner than the one that takes three weeks. A practice running this can honestly tell referrers and clients that it operates around the clock, which is a marketing claim as much as an operational one.
That is the real purchase. The firm is not buying an assistant. It is buying a position against the firm down the road.
Where else this goes
The pattern travels, and the test for it is specific. Look for work where several organisations have to exchange information to finish something, where none of them share a system, where the exchange currently happens by telephone, and where nobody on the call is looking for a relationship.
That describes a great deal of the economy:
- Insurance claims. Adjuster, repair shop, claimant, sometimes a hire car company, all forced to coordinate, none on the same system.
- Property sales. Agent, lender, title company, inspector, surveyor and buyer, working to a deadline on a chain of phone calls.
- Freight and logistics. Shipper, broker, carrier, warehouse, each confirming and re-confirming times that keep changing.
- Specialist medical referrals. The same structure as Wayco's own market, without the legal case attached.
- Building inspections and sign-off. Contractor, subcontractor, inspector and owner, arranging access and approvals.
- Immigration and visa work. Employer, law firm, applicant and a government office, exchanging documents against hard dates.
In every one, the calls are transactional, the coordination is compulsory, and nobody minds who is speaking as long as the thing gets done.
The barrier is the way in, not the technology
It should be said plainly that the technology here is the easy half. Voice models good enough for this work are available to anyone and get better regardless of what any startup does.
What is not available to anyone is a way into a profession. Knowing which records request unlocks which stage of a case, what a clinic will and will not release over the phone, and which firms refer to which providers is local knowledge. And getting a cautious practice to hand over its main phone line is a trust problem, not a software one.
So what limits this whole category is not capability. It is access. Anyone with a genuine way into an industry can build this. Everyone else is holding a component.
Why it is bigger than it sounds
The short description is voice AI for law firms. The insight underneath it removes the hardest problem in the category.
- Nobody on these calls minds that it is an AI. They are transactional exchanges where no one is looking for a relationship, which lifts the constraint that has made voice AI hard everywhere else.
- The coordination is compulsory. The case does not move without it, so the calls happen whether or not anybody buys anything.
- The founder's record is remarkable. A venture-backed company at 13, an acquisition by the United Nations at 14, and a patent at 15.
What to watch
The number that matters is not calls handled. It is the time from first contact to a complete case file, measured against what the same firm managed before. That is the only figure that proves days of telephone tag really did become minutes.
The second is whether Wayco ends up sitting between firms and clinics that did not previously work together. An operator that only speeds up existing relationships is a useful tool. One that starts introducing the parties is something considerably more valuable.