Yardstick Research tear-sheet / healthcare RCM cohort
Akasa
Identity
- Name: AKASA (formerly Alpha Health; rebranded March 2021)
- Founded: 2018 (as Alpha Health) - operating entity reincorporated 2019
- HQ: South San Francisco, California (hybrid hubs in SF, Denver, NYC)
- Funding: $205M+ total raised, most recent round $120M Series C closed June 18, 2024. Anchor investors: Andreessen Horowitz, BOND, Costanoa Ventures.
- Employees: ~235 (Tracxn, Apr 2026); Glassdoor reviews flag cyclical layoffs every 3-6 months
- Leadership: Malinka Walaliyadde (Co-Founder / CEO), Andy Atwal (Co-Founder, VP Engineering), Ben Beadle-Ryby (Co-Founder)
- Archetype: GenAI revenue-cycle automation platform for enterprise health systems, anchored on mid-cycle (coding + CDI) with prebill optimization, claim status, and authorization status as adjacent surfaces
Total score: 77 / 100
(80 raw weighted - 3 pricing-transparency soft penalty)
- Stage fit:
- Foundation (<40 readiness): no - enterprise sales motion, multi-year contracts, integration work; foundation-stage buyers can't onboard
- Pilot (40-59): conditional - yes if pilot is single-product (e.g., Coding Optimizer only) at a $250M+ health system; no for smaller pilots
- Scale (60-79): yes - this is the sweet spot. Cleveland Clinic-scale enterprise rollouts in four months; Cerner-preferred-automation-platform; HITRUST r2 + SOC 2 Type 2
- Optimization (80+): yes - the heritage-RCM-vendor-done-right; integrated Coding+CDI deployment is the differentiator for mature RCM organizations
- One-line verdict: The default GenAI RCM candidate for $250M+ academic medical centers running Epic or Cerner, weakened by undisclosed foundation-model architecture and cyclical layoff reports.
Headline numbers
| Metric | Value | Evidence URL |
|---|---|---|
| Total funding | $205M+ | https://www.crunchbase.com/organization/akasahealth |
| Latest round | $120M Series C, June 18, 2024 | https://www.oreateai.com/blog/akasa-raises-120-million-2024-healthcare-ai/53351832e462e02a0d3814502a595125 |
| Employees | ~235 | https://tracxn.com/d/companies/akasa/__qi5hY-G0qopywHUJR-A5GrvRjRNeAjTtfX6FQ6lbFLw |
| Named enterprise customers | Cleveland Clinic, Duke University Health System, Methodist Health System, University Health | https://hitconsultant.net/2025/10/13/cleveland-clinic-deploys-genai-for-coding-and-cdi-across-enterprise-powered-by-akasa/ |
| Cleveland Clinic deployment | 4-month rollout across all U.S. locations; 100% inpatient coding volume | https://hitconsultant.net/2025/10/13/cleveland-clinic-deploys-genai-for-coding-and-cdi-across-enterprise-powered-by-akasa/ |
| Black Book ranking (2025) | #1 Most Promising Healthcare RCM Startup, 95.2/100 | https://akasa.com/press/akasa-most-promising-healthcare-rcm-startup |
| Compliance | HITRUST CSF r2, SOC 2 Type 2, NIST 800-53, CIS, HIPAA | https://akasa.com/platform/trust-security/ |
| Training corpus claim | 43M+ clinical documents | https://akasa.com/platform/ |
Dimension scores
| Dimension | Score | Weight | Weighted | Evidence |
|---|---|---|---|---|
| AI capability depth | 3/4 | 15 | 11.25 | [VENDOR-CLAIMED + THIRD-PARTY] GenAI is primary value driver; 43M+ clinical doc training corpus; customer-specific fine-tuned models; Cleveland Clinic deployment processed 100% inpatient volume and surfaced missed coding opportunities in ~15% of cases (50% of those accepted by human coders). Capped at 3/4 by [UNKNOWN] foundation-model vendor and pre-2022 heritage. - https://akasa.com/platform/, https://hitconsultant.net/2025/10/13/cleveland-clinic-deploys-genai-for-coding-and-cdi-across-enterprise-powered-by-akasa/ (akasa.md §"AI capability depth") |
| Workflow integration depth | 3/4 | 25 | 18.75 | [VENDOR-CLAIMED + THIRD-PARTY] Cerner-preferred-automation-platform (May 2022) with native automation across eligibility, authorization, claim edits, eligibility denials, claim follow-up. Cleveland Clinic Epic deployment at enterprise scale corroborates working Epic integration. Capped at 3/4 by [UNKNOWN] Meditech / Athenahealth / NextGen connector coverage and no published Epic App Orchard listing. - https://hitconsultant.net/2022/05/05/cerner-akasa-ai-powered-rcm-automation/, https://akasa.com/platform/ (akasa.md §"Workflow integration depth") |
| Vertical specialization | 4/4 | 15 | 15.00 | [VENDOR-CLAIMED + THIRD-PARTY] Pure-play healthcare RCM. No horizontal LLM or non-healthcare surface. In-house RCM operator domain expertise (15+ year average). Black Book #1 most-promising RCM startup ranking validated by 1,303-executive survey. - https://akasa.com/, https://akasa.com/press/akasa-most-promising-healthcare-rcm-startup (akasa.md §"Vertical specialization") |
| Implementation + time-to-value | 3/4 | 10 | 7.50 | [THIRD-PARTY] Cleveland Clinic 4-month enterprise rollout across all U.S. locations is fast for the scale. Capped at 3/4 by [THIRD-PARTY] Methodist Health System's multi-year original (pre-GenAI) journey (2018→2019) showing slower legacy automation cadence. - https://hitconsultant.net/2025/10/13/cleveland-clinic-deploys-genai-for-coding-and-cdi-across-enterprise-powered-by-akasa/, https://www.healthcarefinancenews.com/news/methodist-health-system-and-akasa-partnered-automate-revenue-cycle (akasa.md §"Implementation + time-to-value") |
| Data + compliance posture (HIPAA / HITRUST) | 3/4 | 20 | 15.00 | [VENDOR-CLAIMED] HITRUST CSF Risk-based (r2) 2-year certification (July 2022) + SOC 2 Type 2 + NIST 800-53 + CIS + HIPAA. ZeroTrust VPN + FIPS 140-2 encryption + 24/7 PHI access logging. Capped at 3/4 by [UNKNOWN] AI safety / PHI-training-opt-out default policy, [UNKNOWN] pen-test report publication, and [UNKNOWN] data residency disclosure. - https://akasa.com/platform/trust-security/, https://www.prnewswire.com/news-releases/akasa-achieves-hitrust-certification-reinforcing-commitment-to-healthcare-third-party-privacy-security-and-compliance-301610535.html (akasa.md §"Data + compliance posture") |
| Pricing + scalability | 2/4 | 5 | 2.50 | [VENDOR-CLAIMED + THIRD-PARTY] Scalability proven (Cleveland Clinic enterprise, $205M+ raised, ~235 employees). Pricing transparency penalized: no /pricing page, no published per-claim / per-encounter / per-bed anchor; soft penalty applied per Section 6.7. - https://akasa.com/, https://canvasbusinessmodel.com/blogs/how-it-works/akasa-how-it-works (akasa.md §"Pricing + scalability") |
| Vendor strength + named-customer evidence | 4/4 | 10 | 10.00 | [THIRD-PARTY] $205M+ raised, anchored by Andreessen Horowitz + BOND. Named enterprise customers Cleveland Clinic, Duke University Health System, Methodist Health System, University Health. Black Book #1 RCM startup (95.2/100); top in 17 of 18 categories in 2025 AI-in-Healthcare-Finance review; KLAS Emerging HCIT Companies 2024. - https://akasa.com/press/akasa-most-promising-healthcare-rcm-startup, https://akasa.com/press/top-ai-revenue-cycle-leader-black-book (akasa.md §"Vendor strength + named-customer evidence") |
| Raw weighted total | 100 | 80.00 | ||
| Pricing transparency soft penalty | -3.00 | Per Section 6.7 | ||
| Final score | 77.00 |
Pricing detail
- Public price anchor: None.
/pricingreturns 404. Homepage CTA is "LET'S CHAT." - Inferred model (third-party reporting): Subscription fees + implementation fees + additional services, structured as tiered packages with hybrid software + services + revenue-sharing arrangements creating multi-year contracts. https://canvasbusinessmodel.com/blogs/how-it-works/akasa-how-it-works
- Minimum contract size: [UNKNOWN]
Integrations
- Cerner / Oracle Health: Native - Cerner-preferred automation platform (May 2022). Automates eligibility, authorization, claim edits, eligibility denials, claim follow-up.
- Epic: Demonstrated at Cleveland Clinic enterprise scale, but no published Epic App Orchard listing.
- FHIR / HL7: Standards-based integration claimed; specifics on FHIR R4 / SMART-on-FHIR conformance [UNKNOWN].
- Meditech / Athenahealth / NextGen / Allscripts: [UNKNOWN] no published native connectors; likely built per-customer.
- Clearinghouses / payer connectivity: [UNKNOWN] not publicly enumerated.
Editorial assessment
AKASA is the strongest enterprise-ready GenAI RCM vendor in the healthcare-rcm cohort. The combination of $205M+ committed institutional capital, a Cerner-preferred-automation-platform designation, a four-month Cleveland Clinic enterprise rollout processing 100% of inpatient coding volume, and the Black Book #1 ranking from a 1,303-executive survey is the kind of triangulating evidence that holds up under buyer scrutiny. For a $250M+ academic medical center running Epic or Oracle (Cerner) that wants to extract more value from existing coding and CDI staff without replacing them, AKASA is the default candidate.
The bounded weak spots are real. First, the heritage gap - AKASA is not architecturally an AI-native company. The Prebill Optimization Suite is a Q3 2024 GenAI re-platform layered on a 2018-vintage Alpha Health automation stack. That re-platform is working, but the company carries more legacy infrastructure than peers founded post-LLM, and the ai_native_score should be 75, not the 100 currently in D1. Second, the foundation model is undisclosed - for a product reading PHI and generating billing-affecting output, that's a meaningful gap. Third, cyclical layoffs reported on Glassdoor (every 3-6 months through 2024-2025) is a churn signal absent from vendor-controlled press; a buyer entering a multi-year contract should ask about account-team retention explicitly. Fourth, pricing opacity is real - no published anchor of any kind, soft penalty applied.
Relative to the cohort: AKASA is the heritage-RCM-vendor-done-right. AGS Health is a larger BPO operation with thinner AI. Cohere Health is payor-side and does not overlap. Adonis Intelligence is a smaller, more SMB-friendly competitor with a more AI-native architecture but less enterprise scale. The closest direct competitor is Iodine Software (CDI heritage), which AKASA's Cleveland Clinic CDI deployment displaced. AKASA wins on GenAI depth and integrated Coding+CDI; Iodine wins on CDI-only depth and longer tenure.
Revisit if: (1) AKASA discloses foundation-model architecture and PHI-training-opt-out policy, (2) the cyclical layoff pattern resolves or is publicly addressed, (3) Epic App Orchard listing appears, or (4) a competitor (Iodine, R1, AGS) ships a comparable Coding+CDI GenAI bundle at price-transparent terms.
Right-of-reply gaps
The following [UNKNOWN] items should be sent to AKASA during the factcheck pipeline:
- Foundation model vendor and architecture - Is the GenAI built on OpenAI, Anthropic, Google, or proprietary foundation models? What is the model deprecation / version-management posture? (Affects ai_capability_depth dim and risk profile)
- PHI training-data opt-out default - Is customer PHI included in or excluded from model training by default in the standard BAA / DPA? (Affects data_compliance_posture dim)
- AI safety + hallucination guardrails - What guardrails exist on GenAI output that affects billing? What's the audit log for AI-generated billing recommendations? (Affects data_compliance_posture dim)
- Pen-test report publication cadence - When was the last pen test? Are reports available under NDA? (Affects data_compliance_posture dim)
- Data residency - Where is PHI processed and stored? Multi-region options? (Affects data_compliance_posture dim)
- Epic App Orchard listing - Is AKASA listed in the Epic Showroom / App Orchard, or is integration always custom? (Affects workflow_integration_depth dim)
- Meditech / Athenahealth / NextGen / Allscripts connectors - Native or per-customer build? (Affects workflow_integration_depth dim)
- Minimum contract size / per-claim or per-encounter anchor - What is the entry-level deal size? What's the pricing unit of measure? (Affects pricing_scalability dim and pricing_transparency penalty)
- Account-team retention through 2024-2025 layoffs - How has account-team turnover been managed during the reported restructurings? (Risk signal, not a rubric dim)
- Total customer count + segmentation - Vendor cites "650+ hospitals and 6,500+ outpatient facilities, across all 50 states" - is this all using AKASA actively, or is this aggregate footprint across all years? (Affects vendor_strength_evidence dim)