Irving, TX
AI Lead Qualification Agent for Irving TX Practices
Qualification is a word worth being careful with when the practice in question is a law office or a clinic. In most industries it means deciding who is worth your time. In a professional practice it means something narrower and more defensible: finding out enough to route the inquiry to the right person, quickly, with the facts they need.
Book a Demo
That distinction is not pedantry. It determines whether an automated layer is an asset or a liability, and Irving practices evaluating this should insist on it explicitly rather than assuming it.
Cleod9's AI Voice Concierge asks the screening questions you defined and routes what it collected. It books where your rules allow, captures inquiries with real detail, and warm-transfers to a live person. It does not decide whether you take the matter or accept the patient. A person at the practice does, every time.
Sorting versus deciding
The difference shows up in what happens to somebody whose answers fall outside your usual pattern.
A deciding system tells them you cannot help and ends the call. That is efficient, it is what most people picture when they hear qualification, and in a professional practice it is a mistake with several dimensions.
A sorting system flags the inquiry, routes it, and lets a person look. The caller is told somebody will follow up, which is true, and the practice keeps the option.
Insist on the second. Any vendor describing automated screening should be able to say plainly what happens to an out-of-pattern inquiry, and the correct answer involves a human rather than a rejection message.
Why deciding is commercially wrong, not just risky
Practices are consistently poor at predicting which inquiries become valuable, and a filter built on those predictions removes revenue nobody saw coming.
The person who sounds uncertain about whether they have a real problem is frequently the one with the most substantial matter, because people minimize. The inquiry that does not match your usual practice area is sometimes a referral opportunity worth more than the matter itself. The caller who cannot articulate their situation clearly at eight in the evening is not necessarily the caller who cannot articulate it on Tuesday with a document in front of them.
A screening layer that sorts preserves all of those. A screening layer that decides discards them silently, and the practice never learns what it lost because the call ended before anyone saw it.
What screening questions should actually do
Three jobs, and no more.
Establish who is calling and in what capacity. New or existing, and whether they are calling for themselves or for somebody else. A third-party caller usually cannot answer questions about the person concerned, and pressing produces guesses that end up in your records as facts.
Establish the category. What the inquiry concerns, expressed as something a caller would recognize rather than as your internal taxonomy. Nobody says they have a premises liability matter or a post-operative concern.
Establish urgency as the caller perceives it. Asking how soon they feel they need attention puts the judgment with the person who knows how they feel, and yields a routing signal without the system forming any view of its own.
Four or five questions total. The purpose is to triage and to make the follow-up informed, not to complete the file.
The boundaries, which are not optional
For a medical practice: no assessment of symptoms, no advice about medications, no view on whether something can wait or warrants an emergency room. The greeting opens by telling anyone facing an emergency to hang up and dial 911, before any other question.
For a law office: no legal advice, no opinion on whether there is a case, no fee quotes beyond published consultation pricing, no predictions about outcomes, and nothing implying the firm has taken the matter.
And in a law office specifically, no case substance before conflicts are run. Information a prospective client shares can carry duties even where the firm never takes the matter, so somebody describing a dispute involving one of your existing clients has created a problem in the first ninety seconds. Ask for the practice area as a category and stop.
A question outside the boundary is a handoff trigger rather than something to answer carefully. Have a clinician or a partner review the finished script against these lines. This page is operational guidance rather than legal or clinical advice.
Disclose it, and offer the exit
Say in the first sentence that an assistant is taking the initial details and a member of the practice will follow up. People work it out anyway, and discovering it themselves is worse than being told.
Then make the route to a person immediate when somebody uses it. Three triggers: the caller asks, a rule fires on whatever your practice treats as urgent, or the conversation stops working, meaning the caller repeats themselves or answers a different question than the one asked.
That third trigger prevents most bad interactions and is the one practices routinely forget to configure. Two failed exchanges should route to a person or to a properly captured message rather than to a third attempt.
Be honest about the destination. If nobody answers at nine in the evening, do not build a path that rings out. A message captured well with a callback window you keep beats a phone that tells the caller nobody is there.
What the person receiving it should get
The point of screening is not the screening. It is that whoever picks this up next starts from something.
Call recording runs automatically, so what was said exists as a record rather than as somebody's recollection, and access is governed by the access control list. x-bees is included with Cleod9, and its AI transcription and summaries work on voice calls as well as chat, so the follow-up begins with a readable account rather than a voicemail to replay with a notepad.
The practical effect is that the callback opens by referencing what the person already said instead of asking them to explain again. For a competitive inquiry that is the difference between a follow-up and a cold call to somebody who has since contacted two other practices.
Settle retention deliberately. Records of people who never became clients still contain their information. Ask Cleod9 where recordings and transcripts are stored, what the default retention period is, whether you can set your own, and whether a specific record can be deleted on request. Keep the answers in writing.
Routing rules worth defining
Screening produces a sorted queue, and the sort only helps if somebody acts on it in the right order.
Evening and weekend inquiries first, because those people have had the least opportunity to go elsewhere. Anything flagged on timing next, because a deadline problem does not wait for your workflow. Existing clients and patients with a problem before new routine requests. Everything else after.
Give the queue an owner by role rather than by name, put it in that role's daily routine, and decide what happens when that person is out. A process with a single human dependency fails during the week somebody takes leave, which is exactly when nobody notices.
Where screening goes wrong in practice
Three failure modes, each with an early warning sign worth watching for.
The interrogation. Too many questions, asked too formally, to somebody who is worried. Warning sign: callers abandoning at the same prompt, or a rise in transfer requests early in the call. Cut questions rather than rewording them.
The false negative. A caller routed as low priority who was in fact urgent, usually because the category question captured the wrong thing. Warning sign: somebody in the morning queue who should have been transferred last night. Review these individually rather than statistically, because the pattern is usually a single missing rule.
The sorted queue nobody works. Screening improves the quality of what lands in the list, not the likelihood anyone opens it. Warning sign: capture volume healthy, appointment volume unchanged. That is an ownership problem and no configuration fixes it.
What to measure
- Completion rate: calls ending in a booking, a full capture, or a transfer rather than the caller dropping partway.
- Where callers abandon, by prompt, read from actual recordings rather than a summary.
- Transfer requests as a share of calls, and how early they occur.
- Median time from a captured inquiry to first human contact, split between urgent and routine.
- Consultations or appointments that actually happened, rather than inquiries captured, since the second number flatters and the first pays.
Listen to real calls in the first fortnight. Every practice finds a prompt that fails out loud and a category nobody planned for, and both are quick to fix once heard.
Common questions
Does it turn anyone away?
No. It sorts and routes. Every determination about whether you take a matter or accept a patient is made by a person at the practice.
What happens to an inquiry that does not fit our usual work?
It gets flagged and routed to a person, not rejected. That is the design, and it is worth confirming with any vendor rather than assuming.
Can callers reach a person?
Yes, by asking at any point, and your rules can transfer them without waiting to be asked.
Does it work on our existing number?
Yes. It attaches to the number already on your listings, letterhead and signage.
How many questions should we ask?
Four or five. Enough to triage and to make the callback informed. The complete intake happens with a person afterward and goes faster for it.
Writing the category question well
Of the four or five screening questions, the category question does the most work and is the one most often written badly.
The failure is almost always vocabulary. A practice writes the options in the language it uses internally, because that is the language everyone in the building speaks, and callers then hesitate because none of the options sounds like their situation.
Somebody who slipped on a wet floor at a store does not think of it as premises liability. A patient whose surgical site looks inflamed does not think of it as a post-operative concern. A person whose employer stopped paying them does not think in terms of wage claims. Each of them will pause, ask what that means, or pick the closest option and be routed wrongly.
Write the options as situations rather than as categories, and test them on somebody outside the practice. If a person with no professional background cannot pick an option confidently in three seconds, the wording is the problem rather than the caller.
Keep the list short. Four or five options is comfortable to hear on a phone. Eight is not, and callers will simply take whichever one they can still remember when the list ends.
Overflow during the working day
Practices adopt screening for evenings and weekends, then find the bigger number sits inside business hours.
Every small office has windows where the phones cannot be reached. The opening hour, when overnight messages, walk-ins and calls all arrive together. Lunch, when coverage drops to one person. Staff meetings. Any day somebody is out.
Configured as overflow, the Concierge answers only after a call has rung through unanswered. The caller cannot tell whether the practice is closed or simply busy, and does not need to.
The reason this matters more than the after-hours case is that daytime abandonment leaves no evidence. An evening caller at least reaches voicemail, and a voicemail is a record. A caller who rings during a staff meeting, waits, and hangs up before voicemail has left nothing behind at all. Nobody counts those, which is precisely why they persist.
Keeping the rules current
Screening rules decay in a specific way. They were written to match how the practice worked on a particular day, and the practice keeps changing while the rules do not.
A practitioner leaves and their name is still a routing destination. A practice area is dropped and callers are still offered it. The transfer path points at somebody who moved roles in the spring. None of this announces itself, because the system keeps working exactly as configured.
Two habits handle it. Give the rules an owner by role, with a quarterly review scheduled for the whole year at once rather than whenever somebody remembers. And test the transfer paths monthly by actually calling them, because a destination that rings out is invisible from the administration screen.
Because greetings, questions, hours and transfer destinations are administered by the practice in a browser rather than raised as tickets, the correction takes minutes. The failure was never that changes were hard. It was that nobody owned noticing.
Talking to Cleod9
Cleod9 is a Dallas-Fort Worth provider supporting its own customers, so an Irving practice deals with someone local rather than a distant ticket queue. The platform is described on the Cleod9 services page.
Bring the questions your intake staff already ask a new caller, and be ready to say which of them exist to route the call and which exist to decide something. The first group is what gets automated. The second stays with a person.