Pam review. The AI receptionist that answers dealership calls

The phone is still the main way into a dealership service department.

The service advisor is busy with a customer. The call center works its queue. The call goes to voicemail. Pam takes these calls. It is a voice AI agent, and the company sells it as a staff member, not as software.

I did not work inside the system. Only customers can log in, there is no trial, and the price is not public. I studied the Pam site, the Consumer Reports article, partner press releases, the Pam page on the DealerRefresh forum and an interview with the founder. Every picture below comes from the company site. These are marketing mockups, not live screens.

In short#

• Pam is a US startup with two voice agents for car dealers. The first one answers incoming calls and texts and books service visits. The second one calls the customer base about recalls and lost customers.
• The company works only with car dealers. Partners sell it too. Among them are CallRevu, Tekion, Dealer-FX and the FordDirect program.
• There is one independent review, and it is mixed. Consumer Reports listened to Pam calls. It wrote that the agent is polite, but offers another option every time the customer says no.
• I found no dealer reviews on forums. DealerRefresh has a page for Pam, but no discussions.
• All result numbers are the company's own words. Even these numbers show that night calls give a small share of bookings. The main work happens in the day and with the old customer base.

Where Pam came from#

The legal name is Dream Lab AI Inc., as the footer of the site says. The year 2023 and the town of Sterling, Virginia, appear only in startup databases. The company does not give them. The old domain pamhq.com now sends visitors to pam.ai.

The name comes from Pam Beesly, the receptionist in the TV show "The Office." This works well in the US. The buyer understands at once which job the product does.

There are four founders. They are Samee Khan (CEO), Abdullah Baig (president), Haroun Ansari (CTO) and Omer Zulfiqar (integrations). The company page lists about fifty people. Engineers and salespeople are the biggest groups, about the same size.

In August 2025 the company got new money. Autotech Ventures led the round. HF0, First Harmonic of Ali Rowghani, Idris Mokhtarzada of RocketMoney and Geneva Pacific also took part. The amount was not made public. At that time the company said it worked in 33 US states and three Canadian provinces (Auto Remarketing).

The main screen. An AI workforce, not a program#

The home page title is "The AI Workforce for Automotive."

The subtitle says that dealers hire Pam to book more service and sell more cars. Above the title there is a small badge, "Hired by 800+ Rooftops." A rooftop is one dealership location.

The whole site speaks the language of hiring. Dealers "hire" agents and give them "work." For a manager this makes the budget easier to defend. He compares Pam with the salary of a receptionist, not with a software bill.

Two agents. Incoming and outgoing#

The home page shows two agents. The Receptionist Agent answers people who call in. The Outbound Marketing Agent works the customer lists that the team has no time to call.

Next to them are three numbers. More than 800 dealerships, "20+ years" of call time, and more than $200 million of revenue for dealers. The second number is total call time. It sounds big, but it is about 175,000 hours across 800 stores.

The page for Ford dealers already shows three agents. The new one is Outbound Sales, an agent that makes sales calls. The site gives no details about it.

What Pam does#

• Incoming calls. The agent answers day and night, in English and Spanish. It can be the first step in the phone menu and answer questions that the dealer approved.
• Service booking. It checks free slots in the dealer's own scheduler, avoids double bookings, and sends the customer a text to confirm.
• Sales calls. It finds out what the customer wants. Then it books a visit or passes the call to a salesperson with a short summary.
• Outbound campaigns. It calls and texts customer lists. Recalls, lost customers, declined repairs, retention.
• Manager console. A call log, alerts about dropped or looped calls, and the revenue that the agent brought in.
• Integrations. The site shows the logos of Tekion, CDK, Xtime, myKaarma, Dealer-FX, PBS and Total Customer Connect. In the Raceway Nissan case the company speaks of 14 confirmed integrations.

The console mockup is more honest than you would expect. One call card has this summary. "The assistant repeatedly asked for the caller's phone number." The call did not go further. So the company itself shows that the agent can get stuck. This is useful for a buyer, because such a report helps you find failures.

How Pam sells. Partners instead of ads#

There is no price on the site. Every button leads to a demo request. But Pam has more sales channels than other young companies in this market.

• In January 2025 CallRevu called itself the exclusive Pam partner in the industry (Digital Dealer). CallRevu analyzes dealership calls, and Pam answers them.
• In April 2025 Pam joined the partner program of Tekion and connected to its service scheduler.
• Dealer-FX, a service booking system, named Pam its preferred voice AI partner. But Dealer-FX also added other vendors, for example STELLA.
• Pam became a FordDirect partner. Ford dealers buy digital tools through this program. Podium and Impel are in it too.

For Ford dealers Pam promises "half the cost" for a limited time. The site does not say what the full cost is, or when the offer ends. Launch takes four weeks. Week one is setup, week two is the link to dealer systems, week three is team training. In week four the agent starts to take calls.

The Hyundai case. One month of incoming calls#

On the Receptionist Agent page the company shows a pilot at a Hyundai dealer in Virginia. It does not name the store. In 30 days Pam handled 1,051 calls and booked 579 service visits.

Do the math yourself here. 579 bookings from 1,051 calls is 55%. The 82.7% on the mockup counts only the calls where the customer wanted to book. Both numbers are fair. But compare them with your own numbers on the same base.

The company says Pam saved 69 hours of phone time in a month. This is the time the agent spent on the line. It is about one person working half time.

The Raceway Nissan case. The main company story#

The most detailed case on the site is about Raceway Nissan in Riverside, California. The company says the store was number 74 of 75 Nissan dealers in the state. Only 28% of customers came back for service.

The new general manager, Jason Curnyn, went to service first, not to the sales floor. The dealer used both agents. Incoming calls started in June 2025, outbound calls at the end of July. These are the results, as the company gives them.

The page also has totals for the whole period, and they tell more. The incoming agent took 13,770 calls and booked 1,605 customers. Only 99 of these bookings came after hours, about 6%. So the main value is not at night. The agent takes the day queue when advisors are busy.

The outbound agent worked with 101,662 contacts and reached out almost 440,000 times by voice and text. 26,324 people answered, and 958 booked a visit. This is less than 1% of the list and about 3.6% of those who answered.

For calls to an old customer base, this rate is normal. But the number shows the cost of the method. For one booking the agent makes about 460 calls and texts. A human call center cannot do that much, and the dealer's customers have limited patience.

What customers say#

Pam has one independent review, and it is a rare one. On August 20, 2026, Consumer Reports published an article by Keith Barry, "How Car Dealerships Are Using AI to Upsell Service". The reporter listened to recordings of Pam calls.

The result has two sides. The article says that Pam "was polite but always offered alternatives" when a customer said no. The voice "sounded a little stilted," and once the agent said "Nissan" the wrong way. Pam also cut into the talk to read out a 17-digit VIN. But the article says that the customers on the recordings sounded comfortable.

For a dealer, a persistent agent is a plus. For a buyer who reads Consumer Reports, it is a warning. In the article Pam is an example of a dealer AI that does not take the first no.

The reviews on the company site are full of praise. Richard Bustillo of Murgado Automotive says Pam beat other AI tools and even the group's human agents. Bill Knight of Bill Knight Auto Group says Pam solved an old phone problem.

I found no reviews on G2, Capterra or Reddit. On DealerRefresh the company has a vendor page, and the discussion part says there are no discussions yet.

A problem you should know about. The customer did not know it was AI#

Open sources show no lawsuits, data leaks or big failures. But there is one question that the company itself puts on show. The Pam YouTube channel has a video called "This Customer Had No Idea She Was Talking to AI". The company presents this as a strength.

In the Consumer Reports article the agent also starts the call with the words "Hi, this is Pam" and nothing more. The quotes do not include any words that say it is AI.

For a dealer this is a risk to the store's name. A customer who learns the truth after the call can feel cheated. Rules about telling people they talk to AI differ by country and state. Check them with a lawyer before launch.

Company numbers. What is checked and what is not#

Even Pam's own numbers differ from page to page. The home page says 800+ dealerships. Dealer-FX pages in search results say 400+ and 700+. The 2025 funding news says "hundreds of dealers." Most likely this is growth over a year, but you cannot check it.

What to ask the company before you sign#

1. Price. How much does an agent cost per month, and what changes the price? What is in "half the cost" for Ford, and when does the offer end?
2. Telling the customer. Does the agent say in the first sentence that it is AI? Can I turn this on for my store?
3. Failures. How many calls a month does the agent not finish? Who sees them, and how fast does a person call back?
4. Outbound. How many times does the agent contact one customer, and how does it stop after the customer says no?
5. Cases. Name three dealers on my DMS that I can talk to. How did their numbers change after six months, not after one month?
6. Call recordings. Where do you keep them, who listens to them, and do you use them to train models?

What to take for yourself, even if you do not buy Pam#

Count the night on its own. At Raceway Nissan about 6% of bookings came after hours. Check how many calls you get after closing. If there are few, your main loss is in the day, when advisors are busy.

Work your old customer base. The loudest Pam case did not come from new leads. It came from 14,000 owners who stopped coming. Every dealer has recalls and declined repairs to follow up.

Count conversion in an honest way. Pam shows 82.7% and 55% for the same pilot. Decide in advance which base you use for your booking rate. Then compare vendors on that base.

Present AI as a job. The language of hiring helps you explain to the team why an agent is now in the department. It does not replace an advisor. It is a person on the phone who never goes to lunch.

Videos worth watching#

• Pam. $10M ARR Selling AI to Car Dealerships. Omar Waseem channel, 44 minutes. The founder talks about the company story and about selling AI to dealers. The $10 million yearly revenue is the company's own number.
• This Customer Had No Idea She Was Talking to AI. Pam AI channel, about two minutes. A real call. You can hear how the agent leads the talk.
• AI Employees Are Already Here. ASOTU channel, 16 minutes. A Pam talk at the AutoIndustry.ai summit.

Who Pam fits, and who should look elsewhere#

Pam is strong in one narrow job. The agent answers the phone, books service right into the scheduler and calls the old customer base. The company has partners among the systems that a dealer already uses. So you do not need to change your software to start.

The weak points are clear too. There is no price and no independent dealer review. The only outside review speaks of a pushy voice that is sometimes awkward. The company is proud that a customer did not notice the AI, and the dealer carries that risk.

Look at Pam if your store loses service calls and has a big base of customers who stopped coming. If you need one agent for calls, text chats and the website, look first at Numa and Podium.