This founder is building the back office for American dental insurance

^{A week in AI is like a year in other industries. I hope these issues become your weekly source of AI information, inspiration, and ideas. If we haven’t met before, I’m Amanda Smith. I write about AI and the fascinating folks who are building in this brave new world.}

This week’s founder spent two decades in dental insurance, before building the AI platform that now runs underneath it. His customers’ networks now reach 87% of U.S. dentists. It’s proof that some of the biggest AI opportunities are hiding in an industry’s unglamorous plumbing.

Subscribe | Past essays | Let’s collaborate | Insane Media

🤝 This edition is kindly brought to you by HubSpot for Startups

Logging. Enrichment. Routing. It’s the GTM work that eats a startup’s week. Agents can own it so your team gets back to selling.

On October 28, HubSpot and Dust run Build, Buy, or Hire? live. No panel. We build your first automated workflow in HubSpot and Dust, including a post-call automation and a Slack account ownership layer, and you take the build home.

Join the Live Build

_{We partner with a select group of brands that we use or admire, allowing this newsletter to remain free and independent. Reach out for your campaign here.}

Shaju Puthussery has spent more than 20 years in the dental industry. Seven of those were inside a major dental insurer, moving files by hand and chasing information by fax. 

When he looked back at the payer world he grew up in, the back office had barely moved. Dental insurers still spend nearly 18 cents of every premium dollar on administration, and a routine claim takes 15 to 60 days.

“Most of what happens in that claim is deterministic: two cleanings a year, a $2,000 annual maximum. Rules like that should resolve almost instantly,” Puthussery said. “They don’t, because the data feeding them arrives by fax, on paper, and in records that don’t match. That’s not a hard AI problem. It’s a plumbing problem.”

In early 2022, Puthussery teamed up with Deepak Ramaswamy, who came from quantitative finance, where a trade clears in milliseconds. They started by asking dental insurance executives and provider groups where their operations were breaking down. The answers kept pointing to credentialing, plan configuration and matching claims to the right member. 

The industry loses an estimated $17 billion a year to manual administration.

The founders were sure benefits configuration was the priority. Then, while they were building a claims system for an insurer, its CEO told Puthussery his real problem was onboarding dentists on time. Puthussery had assumed credentialing was solved. 

In fact, it takes 90 to 120 days and costs providers $6,000 to $8,000 a month in lost revenue. The two spent a weekend on it, and Ramaswamy called Monday with a solution.

That work became Lightspun.

Lightspun’s AI has gone through three generations, moving from fine-tuned transformer models to a hybrid with large language models, and now to agentic AI. Rather than chase every new model, Ramaswamy designed the architecture so any model can be swapped out.

Every manually keyed plan rule is a future wrong payment, so Lightspun built benefits configuration first. With clean rules, its newest claims engine decides in a median of 0.3 seconds.

“Out of the box, a model usually gets us to 85 or 90%. Getting from 90 to 98 or 99, which is production-ready, takes people working the edge cases,” Puthussery said. “Humans are part of the automation, not a fallback.”

Lightspun didn’t sign up dentists one at a time. It sold to insurers, dental groups, and third-party administrators. When a large insurer moves its credentialing onto Lightspun, its whole network comes with it. Today, its customers’ networks include more than 175,000 dental providers and reach roughly 87% of practicing U.S. dentists.

Credentialing was the way in. In one demo, Lightspun credentialed 500 providers in under five minutes, and the buyer could do the math on the spot. Because you can’t process a claim without clean provider data, credentialing feeds everything else.

In 2025, the American Dental Association chose Lightspun to power its delegated credentialing service. Smile America Partners uses it to get new dentists treating 500,000 kids sooner.

“In healthcare, the vendor security review is the real gate, not the demo,” Puthussery said. That’s why Lightspun secured SOC 2 Type II, HITRUST and NCQA certifications early.

Next came results buyers could check. Specialty Dental Brands cut onboarding for 300 providers from as long as 12 weeks to under two.

Lightspun also doesn’t ask insurers to rip out legacy systems. Customers prove one module, see the numbers and expand. The biggest hurdle is change management: “A better process lands as a shock to someone who has done the work one way for 15 years.”

Puthussery said the most valuable problems are often the ones people inside an industry file under “that’s just how it works.”

He also says to look at the plumbing underneath the shiny problem. In healthcare, most people chase flashy front-end tools, while provider directories, license verification and plan setup still sit on spreadsheets and faxes. Fix the plumbing every transaction depends on, he says, and you’ve usually found the business.

“The work we went after is boring and safe, and that’s where mistakes happen. If you can automate that kind of work without putting anyone at risk, there’s a real company there.”

AI alone wouldn’t have gotten them in the door, Puthussery said. The certifications did.

If one of these stories stuck with you, I’d love to hear which one.

Speak soon,
Amanda

P.S. If you want to get a founder feature about your own story, reply to this email. If you’d like to reach our newsletter audience (founders, creators, and marketers), click the button below.

If you’re new here, I’m over the moon you’ve joined us! To help me craft content that’s actually useful (and not just noise in your inbox), I’d love it if you took 1 minute to answer this quick survey below. Your insights help shape everything I write.

Dive into our other newsletters - where psychology meets the creator economy, e-commerce marketing, and Human resources.