Healthcare IT · Revenue cycle and claims
Aegis
A three person Y Combinator startup building AI agents that detect denied health insurance claims, draft the appeal, and file it, aiming to turn a two hour manual task into a two minute automated one. Founded in San Francisco in February 2025, it is still pre revenue by any public record and has disclosed no named customers.
A denied claim, then a company
The founders say the company started with a bad bill. Aarav Bajaj had an anesthesia claim denied by his insurer and spent weeks trying to get it overturned, time he felt should have gone to recovery rather than paperwork. According to the founders’ own account on Y Combinator’s site, that experience is what pushed Krishang Todi, Bajaj and Dhanya Shah, three friends from Carnegie Mellon University, to start building Aegis in February 2025.
By June, the company had been accepted into Y Combinator’s first ever Spring batch and was one of ten startups Business Insider picked out of that cohort as a notable agentic AI company. The premise is narrow: watch a healthcare provider’s claims data, catch the denials, and fight them with software instead of a billing clerk’s afternoon.
The problem, in the founders’ own numbers
Aegis frames the opportunity with a specific set of figures, repeated across its Y Combinator launch post and its own site: US healthcare providers lose more than $260 billion a year to denied claims and spend more than $20 billion trying to win them back. Fewer than 15 percent of denials are ever appealed, even though the company says more than half of appeals that are filed succeed. Those are company reported statistics, not figures from an independent study, but they describe a gap that is easy to picture: a billing team facing a stack of denials that would often pay off if fought, choosing to fight only the biggest ones because a single appeal can eat two hours of a clerk’s day.
The founders also point to a second pressure on the same gap. Insurers are increasingly using their own AI to deny claims at volume, which is the argument Aegis makes for why an automated response needs to move at the same speed.
Three founders, one CMU class
Krishang Todi, the CEO, studied economics and math and did fixed income risk modeling at an investment fund in India before Aegis, a background the company connects to the job of ranking which denials are worth fighting. Aarav Bajaj studied computer science and machine learning and worked at Palantir on data driven systems before co-founding Aegis. Dhanya Shah studied information systems and computer science and had built production software at three prior companies. All of this background comes from the company’s own Y Combinator listing and launch materials; none of it has been independently verified against employer records.
Todi has also written under his own name in Forbes Technology Council, a paid contributor program rather than staff reporting, where his August 2025 post argued that medical billing is the next frontier for AI in healthcare.
From idea to demo day
The public timeline is short and comes almost entirely from the company and from Y Combinator. Building started in February 2025. The team joined YC’s Spring 2025 batch, which ran from April to June and was the accelerator’s first ever spring cohort. The batch culminated in Demo Day in June 2025, where Business Insider named Aegis one of ten standout agentic AI companies out of 144 in the cohort, describing it simply as software that helps healthcare providers appeal and win denied claims.
What has happened since Demo Day is described only by the company. Its site advertises integrations with major EHR and practice management systems and payer portals, cites Epic, Athenahealth and eClinicalWorks by name in third party coverage of the product, and says it built HIPAA compliant infrastructure early using compliance tooling from Porter and Oneleet to work toward SOC 2 and HIPAA readiness. None of that has a certificate or filing behind it that this research could find.
What is proven, and what is still claimed
| Evidence | What the record shows | Source type |
|---|---|---|
| Y Combinator batch | Listed as an active Y Combinator company, Spring 2025 batch, founded 2025, San Francisco, team size 3, partner Aaron Epstein. | Independent |
| Demo Day recognition | Named one of “10 of the most exciting AI agent startups” from YC’s Spring 2025 batch by Business Insider, June 11, 2025, which also confirms the founding trio and 2025 founding year. | Independent |
| Team size | Y Combinator’s company page lists team size as 3. A separate blog account describes Aegis as a “four person operation.” | Differs between sources |
| EHR and payer integrations | Described in company materials and a company aligned blog as connecting to Epic, Athenahealth, eClinicalWorks, clearinghouses and payer portals. No integration partnership announcement or customer confirmation found. | Company-stated |
| Performance claims | Company site and materials cite an 80 percent cut in appeal cost, a drop from 2+ hours to under 2 minutes to file an appeal, and a demo dashboard showing a 78 percent overturn rate, $2.4M in tracked revenue at risk and $1.8M recovered in a sample month. These read as product demonstration figures rather than audited results. | Company-stated |
| Compliance posture | Company aligned coverage says Aegis built HIPAA compliant infrastructure from the start and uses Porter and Oneleet to maintain SOC 2 and HIPAA compliance. No public SOC 2 report or attestation letter found. | Company-stated |
| Named customers | None found. The company’s launch post asks for introductions to hospital CFOs and independent practice owners, language consistent with a pre-customer or early-pilot stage in mid-2025. | Not found |
| FDA, NIH, trial and patent records | Searches of FDA 510(k), FDA PMA, ClinicalTrials.gov, NIH RePORTER and NSF award databases for “Aegis” return only unrelated companies: a spine implant maker, a surgical device firm, a drug developer and others with no connection to this company, its founders or its San Francisco address. | Not applicable |
Read plainly, the parts of the Aegis story that outside sources can confirm are small but real: it is a genuine Y Combinator company from the Spring 2025 batch, founded by three named Carnegie Mellon graduates, and it was picked out by a national business publication as one of the batch’s notable agentic AI startups. Almost everything about how well the product actually works, who is paying for it, and what compliance work has been completed, currently rests on the company’s own description of itself.
What to watch
- Whether Aegis names a paying customer, a health system or a billing firm willing to be quoted, rather than describing prospects in the abstract.
- A SEC Form D filed directly by the operating company rather than by an investor vehicle, which would put a number on institutional funding.
- Independent confirmation of the EHR integrations it claims, such as a listing in Epic’s or Athenahealth’s partner marketplaces.
- Publication of an actual SOC 2 report or HIPAA attestation rather than a description of compliance tooling in use.
In their words
“Aegis automates the insurance appeals process end-to-end: from denial detection and compliant appeal generation to submission, tracking, and actionable analytics.”
Krishang Todi, Aarav Bajaj and Dhanya Shah, co-founders, Y Combinator launch post, June 2025 · Company-stated
“That’s when we realized how broken the system was. We started building Aegis in February to fix it.”
Krishang Todi, Aarav Bajaj and Dhanya Shah, co-founders, Y Combinator company page, 2025 · Company-stated
“Providers can focus on healing, not haggling.”
Krishang Todi, Aarav Bajaj and Dhanya Shah, co-founders, Y Combinator company page, 2025 · Company-stated
“Automates the insurance appeals process for healthcare providers to appeal and win denied health insurance claims.”
Samantha Stokes, Business Insider, June 11, 2025 · Independent
“A denied insurance claim is not a ‘no.’ It is a ‘not like this.’”
YesPress staff, San Francisco, July 31, 2026 · Independent
“Narrowness is the strategy.”
YesPress staff, San Francisco, July 31, 2026 · Independent
“Healthcare’s $300 Billion Bottleneck: Why Billing Is The Next Big AI Battleground.”
Krishang Todi, Forbes Technology Council (paid contributor post), August 25, 2025 · Company-stated
Related companies
Sources
- IndependentAegis: AI Agents to win denied health insurance claims
- IndependentThese are 10 of the most exciting AI agent startups to come out of Y Combinator’s first-ever Spring batch
- IndependentAegis (YC X25): The AI Fighting Insurance Denials for Hospitals
- IndependentAegis (YC X25)
- CompanyLaunch YC: Aegis, AI Revenue Recovery Engine for Healthcare Providers
- CompanyAI Denial Management Platform for Healthcare
- CompanyHealthcare’s $300 Billion Bottleneck: Why Billing Is The Next Big AI Battleground
- Public recordFDA 510(k), FDA PMA, ClinicalTrials.gov, NIH RePORTER and NSF award database searches for “Aegis”
Profile researched and written by Healthcare Discovery. Last updated September 29, 2026.
