AKASA

What did the October launch add?

The vendor describes an autonomous mid-cycle platform for inpatient coding and clinical documentation improvement. Its press release says the coding product "reads and reasons across the complete patient record" with evidence behind every code. The homepage describes a custom model for each health system, with the customer setting the autonomy threshold. The release also claims that coding which takes a human 30 to 60 minutes per inpatient encounter finishes in under 90 seconds after discharge.

The wider suite, per the homepage, includes a prebill optimization suite pairing coding and CDI, a CDI assistant, a research assistant and automation for authorization and claim status.

How many hospitals use it?

The homepage says 500 hospitals in all 50 states, representing one in ten US inpatient discharges. The press release adds more than $180 billion in aggregate net patient revenue and inpatient volume up almost 6x in a year. Named customers are Cleveland Clinic, which uses prebill review and "intends to explore" autonomous coding, and Nebraska Methodist Health System. The release quotes its CFO, Jeff Francis, crediting improved revenue integrity, denials, write-offs and payment speed.

Where is AKASA based, and what has it raised?

The press release is datelined South San Francisco, while aggregators say San Francisco. Aggregators give 2018 as the founding year and Alpha Health as the earlier name, and list four co-founders including CEO Malinka Walaliyadde. The About page names no founders and no year.

Totals disagree by source. Aggregators put funding at about $205M across four rounds, including a $60M Series B led by Andreessen Horowitz on March 23, 2021. The $120M Series C is dated June 2024 by some sources and February 2022 by PitchBook, and Tracxn shows only $85M in all.

What backs the accuracy claims?

The vendor says "third-party, blinded evaluations" covering 65% to 85% of inpatient volume found its AI matched or exceeded human coders on MS-DRG, principal diagnosis, quality capture and present-on-admission indicators. Evaluator and sample are not named. KLAS published a short report on April 1, 2026 based on one early adopter, and a KLAS blog says it lacks enough feedback on vendors other than CodaMetrix, Fathom and Nym.

RevCycleAI lists limited independent validation, lengthy implementation, payer-portal fragility and few referenceable clients among AKASA's weaknesses.

What does the platform page state?

It says AKASA integrates with clinical data repositories through APIs, EDI or other means and names no EHR. It claims HIPAA-compliant infrastructure and "SOC 2, NIST-800-53, and CIS certified", without a SOC 2 type. No price is public.

The verdict

AKASA fits a health system with inpatient coding volume that wants autonomous coding configured to its own autonomy threshold. Outpatient facility coding has nothing yet beyond the vendor's "soon".