Why we started Pelica Health
A short founder letter from Catherine and Lalit on why we left Google to build the operating system for value-based care, and what we mean by “operations is the bottleneck, not AI.”
Read the post →Short founder letters and long-form playbooks for the teams that run risk-bearing contracts. V28, RADV, PDC, ECDS, and what we ship in between.
A short founder letter from Catherine and Lalit on why we left Google to build the operating system for value-based care, and what we mean by “operations is the bottleneck, not AI.”
Read the post →A dashboard tells you what to do. An AI agent does it. Why value-based care leaders should stop buying analytics and start buying execution.
A buyer's guide to value-based care software in 2026. How to choose between analytics platforms and AI execution layers, plus a fair vendor comparison.
A 2026 buyer's guide to the best risk adjustment software for IPAs and ACOs: what to look for, a fair vendor comparison, and how to evaluate RADV defensibility.
A fair comparison of Innovaccer, Arcadia, Navina, Reveleer, Pearl Health, and Pelica. The real fork is analytics versus an execution layer.
Navina is prospective and point-of-care; Reveleer is retrospective chart review. Most risk-bearing orgs need both, and one record that closes the loop.
How to choose HEDIS and Stars software for the ECDS transition: pick for ECDS readiness and whether the tool closes gaps or only reports them.
A step-by-step anatomy of how an AI agent closes a care gap: pull context, prioritize, decide, call, follow up, escalate only when needed, and document.
Prospective vs. retrospective risk adjustment: how they differ on timing, data, and risk, and why running both on one record beats two vendors.
How to improve your RAF score under V28 without adding coders. The real levers: accurate suspecting, pre-claim flagging, and correct trumping.
How AI closes HEDIS care gaps faster: prioritize gaps across data sources, route to the right coordinator, run multi-channel outreach, and document the close.
The 2027 Medicare Star Ratings changes to plan for: Tukey, guardrail removal, the equity-reward reversal, cut-point volatility, and measure weighting shifts.
Triple-weighted Part D adherence measures explained: diabetes, hypertension, and statin PDC, the 80% threshold, and the 2026 weighting change.
Member outreach automation across email, voice AI, and SMS, run from one queue so no member gets three calls. What works, what is HIPAA and TCPA compliant.
Turn real-time ADT feeds into a worklist and post-discharge outreach inside the CMS TCM window: 2-day contact, follow-up within 5 days, fewer readmissions.
Yes, healthcare AI can be HIPAA compliant with a BAA, minimum-necessary access controls, SOC 2 Type II, and full audit trails. What each control actually means.
Running value-based care on 8 to 15 vendor portals costs more than the licenses. Duplicate outreach, a BI bottleneck, and retrospective catch-up are the bill.
What a provider network rep walks into a practice with when AI handles the prep. From a 90-minute scramble to a 15-minute agenda with every gap and ADT event.
Why the BI team is a 3-week bottleneck, and how natural-language analytics with provenance changes the shape of the work without replacing BI.
CMS V28 cuts average MA risk scores by roughly 3% as the phase-in completes in 2026. Audit, operationalize, and defend in 90 days.
The three triple-weighted Part D adherence measures (diabetes, RAS, statins) move Stars more than any other lever. The math and the operational levers.
The 2023 RADV rule made findings extrapolated and removed the FFS adjuster. Seven design choices that turn HCC capture into something that survives audit.
NCQA retires hybrid HEDIS by 2029 in favor of ECDS-only measures. Supplemental data volumes increase 35x to 75x per measure. How to prepare now.