Akasa: AI Workflow Automation for Healthcare Revenue Cycle
Akasa sits in a crowded but genuinely underserved segment of the RCM market: the administrative workflow layer. Not clinical coding, not payment posting — but the 40% of revenue cycle staff time spent on eligibility checks, prior auth status calls, claim status inquiries, and denial follow-up. Backed by $85M+ from Andreessen Horowitz and other top-tier VCs, Akasa claims to process 90M+ transactions per month. Here's an honest look at whether the technology delivers and who it's right for.
| Founded | 2019 |
| Headquarters | San Francisco, CA |
| Funding | $85M+ (Andreessen Horowitz, GV, and others) |
| Focus | Eligibility verification, prior authorization, claim status, denial management |
| EHR Integration | Epic, Cerner (Oracle Health), MEDITECH |
| Pricing | % of net revenue recovered or per-transaction model |
| Key Differentiator | Purpose-built for healthcare workflows — not generic RPA adapted post-hoc |
| Competitors | Waystar, Olive AI (defunct), UiPath, nThrive, Experian Health |
What Akasa Actually Does
Akasa's core product is an AI-powered workflow automation platform built specifically for healthcare revenue cycle administrative tasks. The distinction matters: most "RPA for healthcare" tools started as generic automation software (UiPath, Blue Prism, Automation Anywhere) and bolted on healthcare-specific adapters later. Akasa started inside healthcare, which means its AI models are trained on healthcare-specific workflows, payer portal behaviors, and EHR interaction patterns from day one.
The platform targets four primary workflow categories:
- Eligibility verification. Automated real-time checks against payer portals and clearinghouses before service delivery. Akasa can batch-verify large volumes overnight and flag coverage discrepancies before patients arrive.
- Prior authorization. Automated submission and status tracking across payer portals. This is Akasa's highest-value use case — prior auth is deeply labor-intensive, manually error-prone, and disproportionately delays care and revenue. Akasa handles the routine auth requests automatically and escalates complex cases to human staff.
- Claim status inquiry. Automated follow-up on outstanding claims across payer portals. What previously required staff to log into dozens of different payer portals and manually check claim status can now run in the background at scale.
- Denial management. Automated identification and categorization of denial reasons, with intelligent routing for appeal workflows. High-confidence denials that meet programmed criteria can be appealed automatically; borderline cases are surfaced to human reviewers with context pre-populated.
The unifying architecture across all four modules is what Akasa calls its "Unified Automation" approach: a single AI engine that navigates payer portals, EHR interfaces, and clearinghouse connections using a combination of machine learning models trained on the specific behavior of those systems. Rather than building brittle screen-scraping automations that break every time a payer updates their portal, Akasa trains models that are resilient to UI changes — a critical differentiator in an environment where payer portals change constantly.
The Akasa vs. Generic RPA Comparison
Healthcare CIOs and revenue cycle directors evaluating automation tools almost always get pitched by UiPath or Automation Anywhere in addition to Akasa. It's worth understanding what you're actually comparing:
Generic RPA (UiPath, Automation Anywhere, Blue Prism): Rule-based automation that records user interactions and replays them. Extremely powerful for stable, structured workflows with minimal variation. The fatal flaw in healthcare: payer portals are not stable, not structured, and not consistent. When United Healthcare updates their portal — and they do, frequently — your UiPath bot breaks and generates errors until your IT team fixes it. The maintenance burden is constant, and "maintaining the bots" often consumes a larger FTE budget than the humans the bots replaced.
Akasa: AI-native automation designed for the variability inherent in healthcare workflows. Rather than recording and replaying exact click paths, Akasa's models learn to achieve outcomes (get the auth status, check the claim, submit the appeal) across changing portal interfaces. Break rates are significantly lower. The tradeoff: implementation is more involved, pricing is higher, and you're dependent on Akasa's AI models rather than controlling your own automation logic.
For organizations that have tried generic RPA in healthcare RCM and found themselves spending more on bot maintenance than they saved on labor, Akasa's model is compelling. For organizations with small volumes and simple, stable workflows, generic RPA may still be adequate and cheaper.
Akasa occupies the middle ground between pure RPA (brittle, cheap, high-maintenance) and fully autonomous AI coding platforms like Fathom or Nym (which handle clinical documentation, not administrative workflow). If your biggest pain point is the administrative grind — eligibility, auth, claim status, follow-up — Akasa is in the right lane. If it's coding quality, look elsewhere.
What the 90M Transactions/Month Claim Actually Means
📬 RevCycleAI Research
We publish a new vendor deep dive every week.
Independent analysis. No vendor relationships. Get the next one in your inbox.
Akasa prominently cites 90M+ transactions per month. This sounds impressive, but the "transaction" definition matters for evaluation purposes. In RCM automation, a "transaction" can mean many things — an eligibility check, a portal login, a status update pull, an API call to a clearinghouse. The number is large because these workflows naturally generate high transaction volumes at scale: a single prior auth workflow might involve dozens of discrete portal interactions that each count as a "transaction."
The more meaningful metrics when evaluating Akasa are:
- Staff hours reclaimed per 1,000 auth requests. If your team currently spends 15 minutes per prior auth status check, what does Akasa reduce that to? The answer (typically 80-90% reduction for routine checks) is the number that drives the ROI conversation.
- Bot break rate vs. benchmark. How often does the automation fail to complete a task and require human fallback? Akasa's AI-native approach should outperform generic RPA significantly here — ask for client references who can quantify it.
- Denial overturn rate on automated appeals. If Akasa is automatically submitting denial appeals, what's the success rate versus your current manual rate? This is the real revenue impact metric.
Implementation Reality
Akasa's implementation timeline is typically 60–90 days for a standard deployment covering 1–2 automation modules at a single facility. Multi-facility health systems with complex payer mixes should plan for 4–6 months to full deployment and another 30–60 days for the AI models to optimize on your specific workflow patterns.
The EHR integration story is strongest with Epic — Akasa has deep Epic Hyperspace integration and their prior auth workflow is purpose-built for Epic's PACS/auth workflows. Cerner and MEDITECH integrations exist but vary in depth depending on your specific version and configuration. If you're on Epic, implementation is faster and reliability is higher. If you're on an older Cerner or MEDITECH build, get references from comparable environments before committing.
Staff adoption is a real implementation variable. Akasa doesn't replace your RCM team — it changes what your team does. Your eligibility specialists, auth coordinators, and follow-up staff shift from manual portal work to exception handling, escalated cases, and quality review. This is better work, but it requires change management. Organizations that invest in training and change management see faster ROI. Organizations that try to "install and forget" often find they've added technology overhead without changing headcount or workflow sufficiently to capture the savings.
Pricing and ROI
Akasa uses two primary pricing models depending on client size and module selection:
- Percentage of net revenue recovered — typically applied to denial management modules. Akasa earns a percentage of the dollars recovered on automated denial appeals and write-off prevention. Aligns incentives: they only get paid if the automation captures revenue.
- Per-transaction pricing — applied to eligibility, auth, and claim status modules. Flat rate per automated transaction with volume discounts at scale.
There are no public list prices; contracts are negotiated. Typical buyers are mid-to-large health systems with 200+ beds or high-volume ambulatory networks. ROI claims from Akasa marketing are in the 300–500% range over 3 years, primarily through labor cost avoidance (staff redirected from manual portal work) and incremental revenue recovery (fewer missed auths, faster claim resolution).
For evaluation purposes: if you can identify a specific workflow — say, Medicare Advantage prior auth status checks — and you know how many FTE hours your team spends on it today, you can model the per-transaction cost against the labor savings and get a real ROI number. Don't accept marketing-generated ROI estimates. Run the math on your own data.
The Competitive Landscape
Akasa competes in a market that's evolved significantly since their 2019 founding. Key competitors:
- Waystar: Broader RCM platform that acquired Olive AI's automation assets in 2023. Waystar now offers workflow automation as part of a larger claims and payment platform. If you're already on Waystar for clearinghouse, their automation add-on is the path of least resistance — but it lacks Akasa's AI-native depth.
- Experian Health: Strong in eligibility and patient access automation. More enterprise-IT oriented than Akasa, with a broader feature set but potentially more implementation complexity.
- R1 RCM and Ensemble: Full outsourcing players who include workflow automation as part of their services. If you're going end-to-end outsourced, you won't buy Akasa separately. But if you want to automate without outsourcing control, Akasa is the right fit.
- Point solutions (eligibility: Availity, Change Healthcare): For pure eligibility verification, these clearinghouse-integrated tools may be cheaper and simpler. Akasa wins when you want integrated automation across multiple workflow types from a single platform.
Pros & Cons
✓ Strengths
- Purpose-built for healthcare — not generic RPA adapted post-hoc
- AI-native models resilient to payer portal changes
- Strong Epic integration with proven prior auth workflows
- 90M+ transaction scale demonstrates real production deployments
- Performance-aligned pricing on denial recovery module
- Frees RCM staff for higher-value exception work
- A16Z backing signals strong enterprise software credibility
✗ Weaknesses
- Non-Epic EHR integrations (Cerner, MEDITECH) vary in depth
- Higher cost than generic RPA for simple, stable workflows
- Requires change management investment to capture full ROI
- 60–90 day implementation timeline not suitable for urgent needs
- AI model "black box" can make error attribution harder
- Vendor concentration risk as a VC-backed startup (not yet profitable)
- Does not address clinical coding — separate vendor required
Bottom Line
Akasa is a credible, purpose-built solution for the administrative workflow automation problem in healthcare revenue cycle. The A16Z backing, 90M+ transaction scale, and healthcare-native architecture distinguish it from generic RPA adapted for healthcare use. For organizations on Epic managing high volumes of prior auth, eligibility, and claim follow-up work — and experiencing the corresponding labor costs — Akasa offers a real ROI opportunity.
The cautions are practical: non-Epic implementations require more scrutiny, the change management investment is real, and VC-backed startups carry inherent execution risk. Before signing, ask for client references on your EHR, run the ROI math on your own workflow data, and ensure the contract includes data portability provisions if you ever need to exit. But in the AI workflow automation category, Akasa is one of the most serious players in the market.
Akasa, like most well-funded RCM AI startups, is not yet publicly profitable. Their business model requires scale to work — the more transactions they process across more clients, the better their AI models get and the better their unit economics become. This is a reasonable bet, but it means you're partly buying into their growth trajectory. Ensure your contract has reasonable exit provisions if the business changes hands, which is likely within the next 3–5 years given the venture funding lifecycle.
Get every vendor deep dive in your inbox
New profiles drop every Tuesday and Thursday — Waystar, R1, Optum, Olive AI, FinThrive, and 70+ more RCM vendors reviewed.
Vendor marketers read this page. See advertising options →