Introduction
Akasa is a South San Francisco company founded in 2019 that applies generative artificial intelligence to healthcare revenue cycle management — automating and assisting medical coding, clinical documentation integrity (CDI), and related mid-cycle workflows. Malinka Walaliyadde serves as chief executive officer and co-founder, per Akasa’s October 13, 2025 press release announcing expanded partnership with Cleveland Clinic. The company positions itself as a GenAI-native revenue cycle vendor whose models train on each health system’s own clinical and financial data rather than generic public corpora alone.
Cleveland Clinic’s expansion from AI coding to systemwide CDI deployment is Brel’s anchor milestone because CDI requires reviewing documentation, labs, imaging, vitals, and medications — a materially harder multimodal task than code assignment alone. Akasa states Cleveland Clinic was the first organization to complete a CDI pilot and will now implement the solution across all U.S. locations, building on a coding rollout completed in four months earlier in 2025.
What the company does
Akasa’s platform connects documentation and coding into one workflow customized for each health system’s clinical and operational nuances. The AI coding tool processes tens of thousands of encounters at Cleveland Clinic per the October 2025 release. The newer CDI solution gives documentation integrity staff an AI assistant that surfaces key clinical evidence, prioritizes chart review, and supports accuracy across multimodal patient records.
Akasa describes its technology stack as proprietary GenAI combined with deep revenue cycle expertise, powered by custom LLMs trained on institutional data. The about page frames administrative inefficiency as a driver of U.S. healthcare cost burden — a mission statement aimed at CFO and revenue-cycle executives rather than bedside clinicians.
Who it serves
Akasa serves large U.S. health systems and academic medical centers with complex mid-cycle operations. Cleveland Clinic — with Ben Shahshahani, Ph.D., chief AI officer, quoted in the October 2025 release — is the named flagship customer expanding from coding to CDI. Duke University Health System appears alongside Cleveland Clinic on an AHIMA 2025 conference panel cited in the same announcement, signaling peer health-system interest in GenAI mid-cycle lessons.
Buyers evaluate Akasa on whether AI suggestions improve capture of clinical complexity — comorbidities, severity, and procedure detail — without compliance risk from over-coding. Black Book Market Research recognition cited in Akasa press materials targets revenue-cycle procurement committees comparing startup GenAI against incumbent RCM outsourcers.
Mid-cycle staff — coders, CDI specialists, and denials analysts — are Akasa’s daily users rather than bedside clinicians. That user base means adoption metrics focus on charts processed per hour, query response rates, and denial reductions rather than physician NPS scores that dominate ambient AI evaluations.
Company background
Walaliyadde co-founded Akasa in 2019 in South San Francisco with a team blending Silicon Valley engineering and healthcare revenue cycle veterans, per the about page. The company name derives from Sanskrit akasa — sky or ether — framed as a fifth element uniting patients, providers, payers, and pharma in healthcare finance. Akasa operates remote-first with U.S. headquarters in South San Francisco.
By October 2025, Akasa’s Cleveland Clinic relationship had progressed from coding pilot to enterprise CDI — a land-and-expand pattern common in enterprise health IT but still notable in GenAI where hallucination risk makes mid-cycle buyers cautious.
Product and AI capabilities
Akasa trains custom LLMs on health-system-specific clinical and financial datasets, then deploys them as AI assistants embedded in coder and CDI staff workflows. CDI’s multimodal requirements — charts plus labs, imaging, and medication lists — demand retrieval and reasoning beyond single-note summarization. Akasa’s unified platform argument is that coding and CDI improvements compound when both modules share the same underlying GenAI foundation and health-system configuration.
The October 2025 release emphasizes interfaces that surface clinical evidence directly to staff rather than black-box automation — reflecting human-in-the-loop requirements in compliance-sensitive mid-cycle roles where denials and audit exposure follow incorrect AI suggestions.
Key developments
2019: Akasa founded in South San Francisco by Malinka Walaliyadde and co-founders. Early 2025: Cleveland Clinic completes AKASA AI coding rollout across all U.S. locations in four months. October 13, 2025: Expanded partnership deploys GenAI CDI solution systemwide after Cleveland Clinic completed first CDI pilot. October 13, 2025: Cleveland Clinic, Duke, and Akasa panel at AHIMA 2025 on GenAI mid-cycle margin protection.
Why it matters
Akasa matters because revenue cycle GenAI is moving from coding-only pilots into CDI — a workflow where documentation nuance directly affects quality metrics and reimbursement integrity. Cleveland Clinic’s sequential rollout gives other health systems a reference architecture for mid-cycle AI governance. Compare with population-health platforms such as Innovaccer that automate coding downstream of clinical data activation, and ambient vendors such as Ambience Healthcare that generate documentation upstream of coding.
Sector context
Akasa belongs in Brel’s AI in healthcare coverage under revenue-cycle and mid-cycle GenAI, alongside Innovaccer autonomous coding and administrative automation platforms. Ambient clinical AI companies such as Abridge influence documentation quality that CDI and coding AI later consume — complementary layers in the same financial pipeline.
Sources and references
Akasa Cleveland Clinic partnership press release (October 13, 2025) and about page.
- Akasa — Cleveland Clinic CDI expansion (akasa.com, October 13, 2025)
- Akasa — About Us (akasa.com)
Reviewed July 10, 2026. Malinka Walaliyadde is CEO and co-founder. Founded 2019 in South San Francisco.