AI Threat Assessment · 26 May 2026

Toma

Dealership AI
COOKED
6.2/ 10

Toma built something genuinely impressive — a voice AI that can handle 22,000+ dealership calls, save service advisors 30+ hours monthly, and book appointments with the patience of someone who's never had to deal with a 2003 Honda Civic owner asking why their transmission is making 'that noise.' The founders are legitimate AI engineers from Scale, Uber, and Amazon who could have built anything. Instead, they chose to automate the one job in America where being slightly rude to customers is considered a professional qualification.

Business Model
7.5
Automation Risk
4.0
Moat Strength
5.5
Adaptability
6.0
Need Survival
7.0
AI Threat Level
BUSINESS MODEL REPLACEABILITY

They charge SaaS fees to be the phone system middleman between car owners and service bays, which works beautifully until OpenAI's Realtime API makes voice AI a commodity feature that every DMS vendor bolts on natively. The 'automation layer' becomes table stakes, not a standalone product.

7.5
WORKFORCE AUTOMATION RISK

This is the rare case where workforce automation IS the product — they're selling the automation, not getting automated. Service advisors and BDC agents are indeed getting replaced, but Toma is doing the replacing, which buys them a few years of riding the wave before the wave crashes on them too.

4.0
MOAT STRENGTH

Integration depth with DMS systems and years of dealership-specific conversation data create real switching costs, but only until the next AI generation makes their custom training irrelevant. They're building a moat in quicksand — genuine today, gone when GPT-6 arrives pre-trained on automotive workflows.

5.5
AI ADAPTABILITY SIGNALS

Their 'Toma IQ' reinforcement learning sounds impressive until you realize they're building proprietary AI capabilities in the exact same year that foundation models are democratizing those same capabilities. Smart team building the right product at precisely the wrong moment in AI commoditization.

6.0
WILL THE NEED SURVIVE AI?

Dealership phone calls survive, but standalone voice AI vendors don't — the functionality gets absorbed directly into Salesforce Automotive Cloud, DealerSocket, or CDK as a native feature. The need persists; the middleman evaporates.

7.0
Verdict

Toma is automating dealership receptionists just as foundation models are about to automate the automators — they're consultants on the project of their own redundancy. They built a Ferrari to race in a demolition derby: technically brilliant, strategically doomed, and the timing couldn't be more perfectly wrong.

Scores are based on public information and AI analysis. This is an affectionate roast, not a financial assessment. The best companies use this as a mirror, not a verdict.

Roast another →