I used to work as a service advisor - or as the article says, receptionist. This system will not work as described for several reasons.
1. Unless you have a recent job that matches the exact same repair/service, you have incorrectly estimated the cost of the repair. In some states, this matters a lot and will cost the shop money. Unless your LLM only quotes for labor in sane amounts for diagnostic and nothing else, you’re only adding noise. This is a disservice to the client and the shop owner. The client now has an inaccurate quote for work and the shop will get a reputation for being inaccurate in quoting work.
2. Let’s say that you manage to get the exact same job twice. Your machine now needs to source parts. Parts may have been in stock yesterday. The might be out of stock now. If they are in stock, you need to retotal the price since prices are dynamic. Did you teach the agent how to source parts? What rules does it have for sourcing used parts?
3. New jobs can’t be quoted. Even if you taught the machine how to calculate book time and margins, it still has to find the right parts. If your shop does high end work, you know how much of a pain in the ass this is. Also remember that some work requires nonobvious parts - like fluids if you need to remove a part in the way of your goal.
4. The only area I see this being useful in is pickup. The shop can mark a car as done and the LLM can call to inform the client that they can come at a preset, unchanging time to get the vehicle. If the vehicle is staying overnight, the LLM can call with a progress update.
Finally, I’d like to note that this sort of dev work goes beyond hubris. It’s dangerous. The more we assume we know without verifying, the greater the risk. In this case, the dev is risking someone else’s livelihood.
I am certainly not an expert but I agree a lot with your sentiment about the hubris - but the problem as presented in the article makes no sense to me.
If you see a value need for a receptionist, and you suspect that it is costing you thousands of dollars, wouldn't a normal response be, "I should think about hiring someone," rather than turning to an unproven, untested solution like this and leaving your business at the hands of how correct it is? I just cannot understand this line of thinking at all, reaching for a tool that would probably do a worse job than a human would do. Is it not wanting to hire? Not wanting to manage? Hype cycle? Where does this urge come from?
To take this further, if the focus really is the "luxury" part of the market, how do they expect this sort of response to go down well with customers?!
If someone is interested in paying luxury size fees, do they really want some cobbled together chatbot? I say this as an advocate for (high quality) chatbots for various practical needs, but it just seems like it is misunderstanding the customers (or maybe luxury is a bit of a loose term new in the area this mechanic works in?)
These customers own expensive cars - or at least, cars that were expensive when they were new. The car might now be ten years old or more, and the owner bought it used. They want a prestige marque, but the customer does not have the money to buy a new prestige car. So they are looking to save on service.
All the time I see cars with expensive names - BMW, Mercedes Benz - broken down on the side of the road, while old Hondas and Toyotas keep cruising by. Those are the customers for this shop: they spent all their money buying an expensive used car, and now they can't afford to maintain it and fix looming problems; meanwhile the Toyota or Hyundai driver gets maintenance and maybe even takes it to the dealer for it.
A mechanic like this can't afford to hire someone to answer the phone. Such a person is expensive, and these customers want rock-bottom prices despite the car being expensive. So a chatbot is good enough and better than nothing.
The most trustworthy mechanic I used in England had an appointment book pretty much full for four months in advance. He didn't answer the phone, didn't have a computer, just a desk diary. If you wanted him to work on your car you turned up at his workshop and spoke to him. If you were willing to wait until he'd finished whatever thing he was doing he'd take a quick look at your car and suggest a course of action. And despite his full order book if something looked urgent enough and small enough he'd fit you in quite quickly.
He charged reasonable prices, but definitely not rock bottom. He had no need to compete with the bottom feeders because every customer acted as his public relations agent.
Business owners tend to resent having to rely on and pay their workers.
Many of them believe people should line up and volunteer/be forced to work at their companies for free, the fact that they have to pay them is an insult.
They need workers, but workers are not worthy of being needed by them, or paid, so they look for any out at all.
The word you’re looking for is greed. These systems are greed enablers. The narrative used to pump them plays on greed. And so on.
Hiring a person for the job is 3000$ per month? Great let’s try to do this with 500$ and a tangle of vibecoded toothpick bridges!
For a luxury service with generous margins this is a failure-prone mentality.
They'd still try to replace workers, even if their attempted route of automation cost them more than hiring employees would, because of their resentment towards them
Aside from a cost? It's also managing the actual human being, and making sure they have enough work. If the place has 5-10 calls a day, then it's pointless to hire receptionist that will do nothing for 1 hour, and then get 2 minutes chat. It used to be pointless to build software to do that, but since claude code it's cheap enough to make sense.
receptionist as a service has been a thing for like... forever. You are never going to solve the problem of accurately estimating and quoting with AI or an answering service, so pay for someone to answer the phone and take down the details; have a mechanic or trained service rep review and estimate. Cheap code that doesn't solve the problem is not cheap.
Yes, of course. The bot can request information and the customer can provide it if they feel like it, and then someone qualified can call them back when they have their hands free.
But there's no bot, per se, needed at all. An answering machine from 1993 can do this same information-gathering job. :)
So update the device from 1993's new-fangled digital answering machine to 2009's Google Voice, and have it do the transcription from voicemail to text.
Someone will still have to call Bill back about his Honda (which is actually the Kia he bought for his daughter -- Bill is not a very technical guy these days[1] and he confuses such concepts regularly) in order to get any trading of money for services done.
It doesn't take an LLM to get there, and Bill would probably prefer to avoid being frustrated by the bot's insistent nature.
Look, you‘re kicking an open door.
I think LLMs applied like this are just a layer of complexity that os mostly replacing lower level programming solutions that could do the same thing
The transcription + callback loop is honestly underrated.
Most of the value here is just capturing intent accurately
("Honda" vs "Kia" aside) so the mechanic can prioritize
callbacks. A dumb voicemail-to-text pipeline handles that
fine. The LLM layer adds complexity without solving the
actual bottleneck, which is someone qualified picking up
the phone.
But I'm not sure that a bot can be trusted to make good decisions about priority, either. So even if it makes good decisions based on context (which it can increasingly-often do, but does not always do), it lacks the context that is necessary to form the basis of good decisions.
Suppose a message comes into the box with this form: "This is Wendy, can you call me? My car is making that noise again."
The bot might deprioritize that call because it lacks actionable contextual information. "My job as a bot is to get more jobs into the shop. This call does not have enough data to do that, so I'll shove to the bottom of list of callbacks behind more-actionable jobs."
But the mechanic? The mechanic knows Wendy's Ford very well, and he also knows Wendy. She's a been a good customer for over a decade. The mechanic also knows the noise, and that Wendy has 3 little kids and that she's vacationing 900 miles away on a road trip with those kids in that Ford. The context is all there inside of the mechanic's brain to combine and mean that this might be the highest-priority call he gets all week.
Wendy may not have actively relayed any urgency in her message, but the urgency is real and she needs called back right away. She needs answers about what to do (keep driving and look into it when she gets back? pull over immediately and get a tow to a decent local shop? maybe she even needs help finding such a shop?) pretty much immediately. Not because it means more business today, but because it means more business for years.
The mechanic can spot this from a list of transcripts in an instant and give her a ring back Right Now. The bot is NFG at this.
The addition of the bot only adds noise to the process, and that noise only works to Wendy's detriment. When the bot adds detrimental noise to Wendy's situation, it also adds detriment to the shop's longevity.
The presence of the bot -- even as a prioritizing sorting mechanism -- asymptotically shifts the state from an excellent shop that knows their customers very well to a bot-driven customer-averse hellscape.
(And no, the answer isn't to make the bot into an all-knowing oracle that actively gets fed all context. The documentation burden would be more expensive, time-wise (and thus money-wise) than hiring a competent human receptionist who answers the phone, handles the front door traffic, and absorbs context from their surroundings. A person who chatted with Wendy last Thursday right before she left for her trip is always going to be superior to a bot.)
If someone put on their website and voicemail that they were available for calls only from 8-10am (for example), or that they would return my call at that time, I'd make a point to call them then. It's reasonable that people are busy too.
Because the capital owning class in America commonly has an aversion to labor.
Labor is other humans and all their social hierarchy monkey brain bullshit activates in a way that a machine doesn’t. That’s why you’ll see companies spending equivalent or even slightly more money for a tool to do a job over a human being.
Walmart employs this amount of workers only because it is subsided by food stamps and other government assistance. The minute they were forced to actually pay for the labor they employ would fire a lot of people
You are suggesting that if the government gives you a tax break, your boss would lower your salary? Why does your boss wait for the tax break or handout and doesn't just lower your salary now?
Also what's your counterfactual here? If Walmart fired their employees tomorrow and replaced them with robots, those ex-employees would magically no longer need food stamps nor government assistance? (Or more realistically: Walmart could pivot to the Aldi model of labour and replace many low intensity jobs with fewer higher intensity jobs. For the affected workers, the outcome is the same.)
If those ex-workers don't magically get off government assistance, if Walmart is out of the picture, in what sense is Walmart to blame for their poverty?
Conversely: if Walmart laying off these workers would magically improve their welfare, why do these workers wait for Walmart to lay them off?
> Walmart could pivot to the Aldi model of labour and replace many low intensity jobs with fewer higher intensity jobs.
Yes, this is the expected change.
> For the affected workers, the outcome is the same.
No? There are two classes of affected workers:
1. Workers who have been converted to full-time with benefits. These workers benefit from the change.
2. Workers who lose their jobs. These workers are worse off.
Your argument ignores class 1.
I don't think we'll get anywhere debating the relative merits of the tradeoff of those two groups, but I personally prefer the existence of class 1. At least with that class there are some winners.
There's practically no (1). It's a different class of workers, of people than who Walmart currently employs at low intensity and low pay.
People who prefer a higher intensity, higher paying job than the bottom rung at Walmart can already get that kind of job today. They don't need to wait for Walmart to fire everyone else.
Walmart has some of these jobs already, probably. But Aldi and other companies exist. The whole Jeff Bezo's workout at Amazon Warehouses falls in a similar category too: Amazon pays pretty well for the sector and requires no prior experience, but they expect you to stay on your feed throughout.
> Walmart employs this amount of workers only because it is subsided by food stamps
And then those food stamps are used at Walmart, its a win win for Walmart and Walmart. No other country gives their poor food stamps instead of money, I wonder why?
I'm projecting, but I think you're right. Not wanting to manage is probably a large driver. I can imagine that if you've dealt with messy humans before, that a robot receptionist that's not going to show up late, call out when hungover, need an advance for a family member's surgery and then quit, is quite attractive.
Until the robot breaks for reasons unknown and you have to pay for expensive engineering time to fix it. Surprise, since the engineer vibe coded the whole thing, he also has no idea how to fix it except to get the AI to try.
> If you see a value need for a receptionist, and you suspect that it is costing you thousands of dollars, wouldn't a normal response be, "I should think about hiring someone," [...]
If you only have thousands of dollars is savings from the move, hiring someone might be too expensive.
From the post it's clear that the shop has a set schedule of services and prices that the bot is pulling from. All the things you're saying are true for a shop that needs to custom quote each job but do not apply to the situation as presented.
It's clear that the author interpreted the data that way, yes.
And perhaps the shop actually charges the same for brakes whether it is an Ford F150 or a Toyota Corolla.
But that seems very unlikely to me. While they're both very common vehicles, they are also very different and the parts have substantially different costs associated with them.
> Finally, I’d like to note that this sort of dev work goes beyond hubris. It’s dangerous. The more we assume we know without verifying, the greater the risk. In this case, the dev is risking someone else’s livelihood.
I think this is a bit of an overstatement. The dev states it’s her brother’s business, and one can assume he’s asked her to help him out.
Getting the service to be 100% perfect is of course a near impossible challenge, but that’s most likely not the business owner’s concern — they simply want a way to avoid totally losing business. If the service can convert even 10% of customers with a rough quote and timeline it’s most likely useful.
High end shops live off reputation alone. Usually they’re started by a very skilled mechanic who does racing or some other specialty automotive hobby.
The exit plan for these guys is usually to sell the shop. Most buyers are usually skilled white collar workers looking for a new hobby. The shop folds after that because they no longer have the same connections to the specialty community.
You can get business outside of the specialty auto scene. In fact, it’s required since that’s what actually makes money. Google reviews and word of mouth are king here.
So do you remove the owner from the customer experience? I wouldn’t. But if you are going to do that, then, understanding the risk is important.
Right it's almost like the preferred solution – since the guy is the source of knowledge, but is under the hood – would be just installing a phone with a loud speaker in a garage that he could answer without using his hands.
He knows the prices and the parts, there isn't enough calls to hire a receptionist, and voice controlled systems are quite easy to make. If you use local voice recognition model, you could even mention neural networks in a write up, it's a win-win
And it's quite cheaper too, I'd estimate around 200$-300$ for a room. Most of it for a good microphone.
There's a lot of stupid keyboard punching that you need to do to create a proper quote. You can't do that from under a car.
Someone comes in for a timing belt on a car that needs timing belts. Ok, you've got the parts and labor on the belt itself, those are easy to just look up from a table. But, do you just replace the belt, replace other things related to it as well, or perhaps even go above and beyond and proactively do a water pump or something while you're in there. That's gonna depend on the relative cost/labor differences of those, the customer's intentions, the value of the car, etc.
What you really need is someone to answer the phone, pull all that info and then say "hey Jim there's some guy on the phone who needs X, the options are" and then run down all the reasonable combinations and let the tech exercise judgement. This is basically what the service writer does but ideally they get good and don't need to bother the tech for "minor" judgement stuff.
And that's all assuming that this is a cookie cutter job.
The shop can mark a car as done and the LLM can call to inform the client that they can come at a preset, unchanging time to get the vehicle. If the vehicle is staying overnight, the LLM can call with a progress update.
This doesn't need an LLM. It can be and has already been done for many years with a simple TTS.
The solution is to have the quote manually prepared, entered into the system of record, and then automate the outbound phone call to let the customer know, and agree to the work based on the accurate quote. That's the savings.
One of my specialties is creating call centers using Amazon Connect. I agree with everything you said. I wouldn’t let an LLM go near doing a quote or ordering parts.
I share your skepticism but it does seem like the author addressed 1 and 3:
> When a caller asks something that isn’t in the knowledge base, the AI doesn’t guess. It tells the caller it doesn’t have that information, asks for their name and a good callback number, and saves that to MongoDB. Dane gets a list of callbacks to return — no lost leads.
> The escalation path is not an edge case — it’s a core feature.
I haven't been a service advisor before, but if it's anything like working the phones at other retailers, you get a lot of the same questions over and over again, and a bot could certainly answer those things correctly.
> The phone rings, he can’t answer, the customer hangs up and calls someone else. That’s a lost job — sometimes a $450 brake service, sometimes a $2,000 engine repair — just gone because no one picked up.
Sure, that's a problem, but...
> Dane gets a list of callbacks to return — no lost leads.
Yeah. So. I'm still going to hang up, phone somewhere else, and you get no business. I'm also doubly annoyed because not only did you waste my time speaking to a computer, it couldn't answer the question so I'm now worse off than if you'd ignored the call.
Yeah - this scenario presupposes that if I need my car fixed I'm going to wait for you to give me a call back, rather than continue working down my list.
The AI doesn’t have to solve every problem to solve some problems. If it can answer 10% of questions, isn’t that 10% better than having all of them go to voicemail unanswered?
The data the bot has to work with is stated to already be available the website.
Therefore, I'd never call on the phone to find those answers -- but those are the only answers the bot has to offer.
The only reason I'd ever call is for answers that the website (and therefore, the bot) does not provide. Calling on the phone and getting a bot that insists on giving me data that I already have would only serve to waste my time and frustrate me.
It would probably frustrate me enough to hang up and call a different shop immediately, and name-and-shame the place.
I know how to Google shit. By the time I start dialing telephone numbers, I've already Googled this shit.
When I call a local shop I want to talk to someone at that local shop (or at very least, their voicemail) -- not a regurgitating bot.
But, again, that's just me.
---
So I'm imagining my dad, who's in his mid-70s and has never Googled a single thing in his entire life. At least superficially, he sounds like an ideal candidate that can be helped with this automated receptionist.
Except: When he calls the shop and has to talk to the bot instead of a person or their voicemail, he's also definitely hanging up immediately and calling the next place on his list. This doesn't help him at all, nor does it help the shop.
---
For the shop, the cost of frustrated people who vent to their friends about the experience may very well be higher the cost of not always being available to answer the phone.
There is no way to ensure that the AI doesn't guess. You can do all the prompting and RAG you want but sometimes it's just going to make shit up and ignore instructions
It might be the same calls if you work at a shop that only works on one model year vehicle doing a set of services. Note that services are not repairs.
If you are already drowning in work, make everything expensive enough to cover the errors. You always have some % chance to lose money with each customer. If the estimate is to high one can give a discount afterwards.
Could play the call on the speaker and decide if it is worth dropping your tools and walking to the phone. Those 10-20 daily marketeers are definitely not it.
> Finally, I’d like to note that this sort of dev work goes beyond hubris. It’s dangerous. The more we assume we know without verifying, the greater the risk. In this case, the dev is risking someone else’s livelihood.
Every single phone call to a business will be having this in the back office within 5 to 10 years.
Couldn't you blast through all of that with some kind of warning like: "All labor and part prices are estimates, for exact pricing leave your name and number and we will get back to you".
From the Washington state attorney general’s website:
“ Estimate: You are entitled to a written price estimate for the repairs you have authorized before the work is performed, only if you deal face-to-face with the facility and the work is expected to cost more than $100. Once you receive an estimate, the facility may not charge you more than 10% above the estimated costs without your prior approval.
The estimate includes, among other things: the odometer reading; a description of the problem or the specific repair requested; choice of alternatives for the customer; the estimated cost; labor and parts necessary for the specific diagnosis/repair requested”
So the LLM builds an estimate. Maybe it’s under 10% difference when the customer walks through the door.
When it’s not, there’s a big problem. Yes, this is still before work has begun, but now you’ve wasted the customers time. And potentially wasted their money if the vehicle was towed in.
> “ Estimate: You are entitled to a written price estimate for the repairs you have authorized before the work is performed, only if you deal face-to-face with the facility and the work is expected to cost more than $100. Once you receive an estimate, the facility may not charge you more than 10% above the estimated costs without your prior approval.
I don't see how "estimates" given over the phone by the LLM and "estimate" as mentioned in this quote refers to the same thing, for the legal purpose of this statement. This would be strictly before repairs have been authorized, and it's obviously not a written estimate. If the client requests a written estimate, it would have to come at a later time after the human mechanic reviews related costs (like specialty parts availability/ship times), or the client bringing the machine in for physical inspection by the mechanic.
From my understanding of the article, it doesn't sound like the LLM is built to fully circumvent a customer phone call by the owner/mechanic before approving a job request unmanned: It's simply to not let go of a client lead because there was no one available to answer the phone, without needing to hire a full-time phone receptionist.
It seems highly unlikely a customer is towing their vehicle in without talking to the mechanic directly first, who now has some context and the ability to sift nonsense requests from realistic ones from the logs before calling or writing to the customer on their own time with all the expert nuance necessary.
I didn’t say it was a written estimate. I said the opposite.
Do you know how towing a car into this particular shop works? If so, please enlighten me.
In most shops, mechanics do not talk to customers. Mechanics get paid to work on vehicles - not talk on phones.
Regardless, the potential for sticker shock exists if the LLM and the mechanic disagree on pricing. You can and will lose customers due to this. I’ve seen it happen. That’s why service advisors are trained to only quote for diagnostics over the phone.
Finally, in the sales training we got, we were taught to not compete on price. This rule doubly applies to a high end shop. They make their money by competing on quality and timeliness. Adding the LLM to the equation compromises both of those.
I didn't mean "estimate" in some technical/legal sense, replace the word with "our best guess" or whatever terminology is used by the 1000s of companies that are using LLMs but not being held legally responsible for what they say.
Realistically, every single mechanic has Google reviews that say “they said it would be $C but it was $D” and so on. You’re claiming a level of rigor I don’t think any real mechanic has.
I'd like to point to tabs in browsers as a problem area. Ctrl + T is a problem.
I'll set the scene that I think most of us have experienced: you're working on a project. You start down the rabbit hole of research to find a solution to something. Maybe you find it quickly somehow. But this case, you don't. The problem is too big for an easy answer and instead requires synthesis and reflection.
Eventually, after opening 50 tabs and only closing the immediately useless stuff, you find that you need to circle back up the problem solving chain. The problem is that you have 45 tabs open and no method to the madness that is clearly visible.
This further compounds if you're trying to solve a new problem with an existing set of tabs that haven't been cleaned out from the last problem.
Nowhere in this process is the UX leading you to solving a problem.
My half-baked solution is to allow for the user to enter "research mode". When a new tab is opened, the browser halts the user and prompts for what they found on the last tab that led them to opening this new tab. When the user leaves research mode, any leafs left should also prompt for a summary or omitted as irrelevant. Then, once all the tabs have been accounted for, a report can be generated which shows all the URLs and the user's notes. Bonus points if allows generation of MLA / APA citations automagically. Further bonus points if I can highlight sections of text / images while in research mode to fill my new tab questionnaire as I go.
1. Unless you have a recent job that matches the exact same repair/service, you have incorrectly estimated the cost of the repair. In some states, this matters a lot and will cost the shop money. Unless your LLM only quotes for labor in sane amounts for diagnostic and nothing else, you’re only adding noise. This is a disservice to the client and the shop owner. The client now has an inaccurate quote for work and the shop will get a reputation for being inaccurate in quoting work.
2. Let’s say that you manage to get the exact same job twice. Your machine now needs to source parts. Parts may have been in stock yesterday. The might be out of stock now. If they are in stock, you need to retotal the price since prices are dynamic. Did you teach the agent how to source parts? What rules does it have for sourcing used parts?
3. New jobs can’t be quoted. Even if you taught the machine how to calculate book time and margins, it still has to find the right parts. If your shop does high end work, you know how much of a pain in the ass this is. Also remember that some work requires nonobvious parts - like fluids if you need to remove a part in the way of your goal.
4. The only area I see this being useful in is pickup. The shop can mark a car as done and the LLM can call to inform the client that they can come at a preset, unchanging time to get the vehicle. If the vehicle is staying overnight, the LLM can call with a progress update.
Finally, I’d like to note that this sort of dev work goes beyond hubris. It’s dangerous. The more we assume we know without verifying, the greater the risk. In this case, the dev is risking someone else’s livelihood.
reply