HomeCompaniesPelica

Transforming healthcare operations with AI agents

Pelica is the AI-native operating system for value-based care organizations. We unify claims, EHR, pharmacy, lab, and ADT data into one live record and put an AI copilot next to every team that depends on it: quality/Stars, pharmacy, care management, provider network, and risk adjustment. The copilots don't just surface insights. They do the work, from engaging members for preventative care using multi-channel communications to preparing provider field reps with everything they need before they walk in the door. So teams spend time on relationships, not admin.
Active Founders
Lalit Kundu
Lalit Kundu
Founder
building healthcare's AI operating system Previously grew to Staff Engineer at Google within 6 years, where I led a 45+ engineer org to deliver $XX-B ARR solutions. Before that, I built core payments infrastructure for Google Pay India. Programmer at heart, ACM-ICPC World Finalist and coach for India’s IOI team (8 medals).
Catherine Zhao
Catherine Zhao
Founder
Building healthcare's AI operating system. Prev Google APM & SWE, building YouTube, ChromeOS, image & video GenAI infra, and enterprise AI agents. Dartmouth CS.
Company Launches
Pelica: the AI operating system for value-based care
See original launch post

Hey everyone 👋 We’re Lalit and Catherine, co-founders of Pelica Health.

TL;DR

Pelica is the AI operating system for value-based care. We pull claims, EHR, pharmacy, lab, and hospital feeds into one canonical record, then run AI copilots that read and write to it: risk adjustment, Stars and quality, pharmacy, provider network, and care management, plus an AI data analyst you can ask anything in plain English. The copilots do the operational work themselves: capturing HCCs in real time, closing gaps once across every payer, and making outreach calls in the member’s language. HIPAA-compliant, SOC 2 Type II, and RADV-defensible.

We’re live with a physician-led IPA in New York, one of the largest independent medical groups in the Northeast: 2,500+ hours saved and higher CMS Star Ratings so far.

▶️ Watch their story: https://youtu.be/2CGmyif9Yd4

uploaded image

An AI agent works one gap end to end: provider office, pharmacy, the member in their language, the payer portal, then a live call. Closed in 3 minutes of staff time.

🚩 The problem

Value-based care pays a medical group or health plan to keep a whole population healthy, and billions in CMS bonuses ride on it. The teams who own those contracts decide whether the org makes money, and they cannot operate:

• Their data lives in 8 to 15 systems. By the time a problem surfaces, the submission window has closed.

• Risk calls a member. Quality calls the same member. The practice calls again, sometimes in a language the member doesn’t speak.

• The best predictive model is worthless if it never reaches the provider.

• The rules keep moving. Hybrid HEDIS is retiring into ECDS volumes many times higher per measure, and the V28 phase-in costs the average MA plan about 3% of RAF.

Every new gap gets a new point vendor, each with its own portal, data model, and login. These teams do not need another point tool. They need an operating system.

🛠️ What we built

One canonical record, with an AI workforce next to every team. Six copilots:

Risk Adjustment: real-time HCC capture with trumping logic and RADV-defensible chain of custody on every code.

Quality and Stars: gap closure unified across every payer, with year-round forecasting for the post-hybrid HEDIS world.

Pharmacy and Part D: PDC lift on the three triple-weighted measures that move Stars the most.

Provider Network: auto-built practice agendas that cut meeting prep from 90 minutes to 15.

Care Management: ADT-driven transitions-of-care worklists, with no duplicate calls across teams.

AI Data Analyst: plain-English questions against the record, in place of the BI ticket queue.

uploaded image

Risk Adjustment: the agent works the EHR and payer portals like a person and captures a defensible HCC.

uploaded image

AI Data Analyst: ask your RAF in plain English, get an answer and a chart in seconds.

uploaded image

Provider Network: reps walk in prepped, and the visit is captured into tasks automatically.

It runs on the systems VBC teams already use: Epic, Cerner, Athena, eCW, payer SFTP, PBM feeds, ADT, and labs, with SMART-on-FHIR overlays in Epic and Athena. A new source onboards in days, and the copilots start working the day after.

📈 Early proof

That physician-led IPA went from three uncoordinated calls per member to one outreach, in the right language, that closed three gaps at once. No new headcount. One of their leaders put it simply: “You don’t just get software, you get partners.”

👫 Why us

Catherine and I met at YouTube and have built large-scale systems together since. We saw the same gap from two sides: clinical AI that could not operate, and operations software that could not reason clinically.

Catherine led AI work on Gemini and enterprise agents at Google, did clinical AI research at Stanford School of Medicine, and saw front-line care at clinics in California and New Hampshire. BA in Computer Science from Dartmouth.

Lalit spent 8 years building software for regulated industries at Google and YouTube, most recently as engineering and ML lead on YouTube Payments, running a 45-person team on products driving $16B+ in revenue. Wharton MBA, ACM-ICPC World Finalist.

Backed by Y Combinator, SurgePoint Capital, Upside Partnership, ISEED Ventures, and Progressive Ventures.

🙏 Our ask

• Warm intros to VPs of Risk Adjustment, Stars, Quality, or Pharmacy at MA plans, IPAs, MSOs, or ACOs.

• Intros to ACO executive directors running MSSP or REACH.

• If you run risk-bearing contracts at a health plan, IPA, MSO, ACO, or medical group, email lalit@pelica.com or book a demo at cal.com/pelica/pelica.

Follow us on LinkedIn and X, and learn more at pelica.com.

Previous Launches
Transform Value Based Care with AI Agents
YC Photos
Jobs at Pelica
San Francisco, CA, US / Remote
$80K - $150K
3+ years
San Francisco, CA, US / Remote
$80K - $150K
1+ years
Pelica
Founded:2025
Batch:Spring 2025
Team Size:5
Status:
Active
Location:San Francisco
Primary Partner:Nicolas Dessaigne