What does Innovaccer actually do? A look under the hood

Today I’m chatting with Abhinav Shashank, CEO at Innovaccer. Innovaccer is a company you’ve probably heard of but thought “wtf does that company do” or “wait is that just innovation and accelerator as a combined word?”

Today we will:

• Explain what they do with case studies and screenshots
• Talk about how they differ from EHRs and the solutions EHRs offer
• Understand how they work with third-party applications vs. build their own (and even which apps they’ve tried that haven’t worked out”
• What customers are asking them in terms of AI applications

This is a sponsored post - you can read more about my rules/thoughts on sponsored posts here. If you’re interested in having a sponsored post done, let us know here.

—

NK: Pretend I’m a small child that just gained object permanence and the ability to understand basic concepts in healthcare. Can you give me the easy to understand version of what Innovaccer does? 

Abhinav: Innovaccer pulls together a hospital or health system’s scattered data—clinical records, claims, scheduling, and other operational feeds—into one place so it's clean and standardized, and linked together. 

This might sound easy, but it’s taken us seven years and a lot of upfront capital to create a system that does this. Do you know how hard it is to follow one patient through a large hospital? It’s shockingly hard. We basically built an entire engine that can ingest data from virtually any source within a hospital, harmonize it to common data ontologies and create a patient index so all of these different sources of data can link up and update in real-time when changes are made. Our first deployment cost us $12M+ to do this harmonization for a relatively small contract, but every deployment gets easier as we learn the shape of different databases and feeds.

You can almost think of this as healthcare "intra-operability" - the ability to get the data you need within your own organization. That unified dataset sits on top of existing electronic health records (EHRs) + other databases, filling in gaps. The goal is to make that data queryable, up-to-date, and ready for applications to be built on top of it.

Once the data foundation is in place, Innovaccer layers on purpose-built apps, agents, and copilots for common tasks: 

• population-health dashboards that flag high-risk patients
• tools that help clinicians draft visit notes and spot missing diagnoses
• scheduling software that matches patients to the right slot, etc.
  

Because every tool draws from (and writes back to) the same data hub, information flows automatically between departments, letting marketers, call-center agents, clinicians, and care managers all act on the same facts without extra hand-offs.

[NK note: it was interesting to watch the first half of this video on Kaiser’s request for proposal for a population health/analytics solution. It walks through the thought process and steps that go into an RFP, how they choose to compare different solutions, etc. They end up picking Innovaccer here, but this helped me understand how the landscape of solutions here work AND what a company factors into their decision when picking a vendor]

NK: I know you work with providers, payers, governments, and life sciences companies. Can you give me an example of what issue each of these entities might come to you with, and what you would do to solve it for them?

Abhinav: 

Life sciences 

Problem: A life science organization launched a new therapy for Multiple Myeloma, but wants evidence on how the therapy performed in real world clinical practice with Medicare patients. This makes it difficult to demonstrate effectiveness in payer negotiations and understand the patient journey outside of controlled clinical trials.

Solution: Innovaccer provides access to Medicare and Medicaid claims datasets, analytics, and actuarial expertise, allowing life sciences organizations to create a complete picture of therapy utilization patterns that support research publications, strengthen their payer negotiations, and ultimately expand their research across multiple products. 

Provider

Problem: A health system’s ability to succeed in value-based care agreements is challenging. Many provider organizations are still new to population health management, and cannot operate these programs without technology and expertise. They may have a large affiliate provider network that participates with the ACO but it may be difficult to manage their performance, and we’re not achieving the criteria that we negotiated with payers.

Solution: Innovaccer:

1) aggregates the organization’s clinical data across different systems and link it to claims data

2) provides population health analytics and workflow products that makes it easy for providers and care managers to close gaps in care and proactively engage high risk patients

3) tracks operational performance across a network of providers ensuring that they are on track to meet payer obligations

4) builds and simulates financial scenarios to strengthen contract negotiation with payers, enabling better health outcomes for patients and stronger financial performance.

You can also see our referral copilot here, if you’re interested.

Payer

Problem: A large payer is struggling with fragmented care coordination across multiple settings (office, pharmacy, home), inefficient and expensive workflows for chart retrieval, and poor member engagement hurting HEDIS and Stars performance. 

[NK note: These are bonus payments that health insurance gets for hitting certain quality metrics. Just don’t ask how they decide what “quality” is, you’ll get disappeared.]

Solution:
At pharmacies, we give prompts and rewards to pharmacists to:

• Make sure patients stay on key meds (diabetes, blood-pressure, cholesterol).
• Offer vaccines and quick screenings (e.g., flu shot, breast- and colon-cancer tests).

These actions lift pharmacy-driven Star and PQA scores.

At physician offices, we show providers during the visit if there are any missing tests or documents—blood-pressure follow-ups, annual wellness visits, depression screenings, etc.—so they can close those gaps on the spot. This is via our InNote app, which slides over the EHR interface and shows the top priorities for each patient encounter.

Government: 

• Problem: A county health department is struggling to incorporate non-traditional data sources into their care delivery ecosystem for underserved populations, like social services data or data from criminal justice systems. Without cross-sector data sharing across health departments, social services, and criminal justice agencies it becomes difficult to direct citizens to the right resources and programs for them.
• Solution: We deploy a unified data platform that integrates all of their disparate data sources into a single record for each citizen. We provide the analytics needed for population health management and health equity analytics to identify and address disparities across the county. We power care management workflows for community health workers to proactively engage the most vulnerable patients.

NK: Let’s say I’m a new customer, just signed the Innovaccer papers that took 12 months to get through procurement. What happens next? Can you walk me through the process?

[NK note: The implementation nerds are really going to get a kick out of this one, I did it for you. Everyone else can scroll past, enjoy the air outside.]

Abhinav: This depends a lot on what is purchased, but providing a very basic view of a typical platform deployment below.

1. Welcome & Kickoff:
   
   1. Within the first week or two, you’re introduced to your Customer Success Director and Customer Engineering Manager.
   2. A formal kickoff call is scheduled with your executive stakeholders to align on goals, timelines, and priorities
2. Discovery & Readiness Assessment
   
   1. Deep-dive discovery sessions are held with clinical, IT, analytics, and operational teams.
   2. Innovaccer assesses electronic health records, claims data sources, ADT feeds, and social determinants of health needs.
   3. Security and compliance requirements are reviewed.
   4. Deliverables are:
      
      1. Project Charter
      2. K03 Document (a big excel workbook for requirement gathering)
      3. Organizational Hierarchy
      4. Attribution Methodology
3. Data Integration & Pipeline Build
   
   1. Customer Engineering Manager’s team builds data pipelines to ingest and normalize data from EHRs, HIEs, claims, etc.
   2. Data QA is completed - confirmation of required data variables is done at this time
   3. Provider Rosters are shared
   4. Innovaccer’s Data Activation Platform maps and matches data into longitudinal patient records.
4. Clinical Application and Workflow Design
   
   1. Clinical SMEs work to gather requirements on the clinical workflows that will be built within the platform
   2. Confirmation of reporting requirements are captured
5. Solution Configuration
   
   1. Care Management, Analytics, Provider InNote, or other modules are configured based on priorities.
   2. User personas (e.g., care manager, medical director) are defined.
   3. Role-based workflows, rules, and reports are tailored.
6. Clinical Validation & UAT
   
   1. Clinical leaders and end users test workflows in staging environment (goal is sign off)
   2. Validation ensures that the logic, data, and workflows are aligned with SHN’s goals.
7. Training & Go-Live
   
   1. A tailored training program is delivered — virtual or onsite — using live demos, playbooks, and quick-start guides.
   2. A formal go-live date is confirmed and supported with hypercare.
8. Adoption & Value Realization
   
   1. Weekly Active Users (WAU) and other engagement metrics are tracked.
   2. Innovaccer provides usage dashboards and clinical performance reports.
   3. Regular meetings drive optimization and expansion discussions.
      
      1. Quarterly Business Review scheduled

NK: The EHRs themselves have started building more of their own applications  (e.g. Epic has HealthyPlanet). How do you compete/differentiate with that?

Abhinav: 

• EHRs today are like Microsoft was in the early 2000s. We eventually reached a point where Microsoft was no longer the only lens through which people accessed information. Google came along and became the system of intelligence—how we search and act on information. Both coexisted, but served fundamentally different functions. That’s the best way to think about us. EHRs are systems of record. Innovaccer is building the system of intelligence—one that sits across fragmented data sources and helps healthcare teams reason over it, take action, and automate workflows.
• One core area of differentiation is that healthcare networks are complex, and even large health systems using Epic or Oracle Cerner tend to have lots of physicians using a wide variety of EHRs. No single EHR application can provide consistency across this type of diverse ecosystem. As the Chief Population Health Officer or Chief Marketing Officer of a large health system, it becomes near impossible to roll out new strategies or programs across the hospital network if every region or ministry or state operates with completely different EHRs and data infrastructure. 
• At the risk of stating the obvious, not all of the EHR applications are actually top notch. For example, top EHRs offer solutions for marketers to manage the patient experience and many of these products are poorly reviewed. Health systems may accept using a worse product offered by their EHR vendor to achieve deeper integration, but we aim to provide deep enough integration that health systems won't feel that they have to use an inferior application. 
• An EHR deployment at the heart of a large health system may have pretty comprehensive data coverage, but smaller EHR deployments at affiliated or independent practices often will not. The most fantastic application from a user experience standpoint still doesn't do you any good if it's relying on incomplete or outdated data.
• Even when they have access to more comprehensive external data, EHRs have a tendency to overemphasize their own "internal" data. As an example of what I mean, I recently switched health systems. When I showed up for my first visit at the new place - where I'd last been seen maybe 10-11 years ago -  the meds list that was presented to the MA rooming me was from 10-11 years ago.  There was one med that I literally had no idea what it was for, although presumably it was prescribed to me at some point.
• You might say, ok, that's not ideal but it's understandable, EXCEPT they DID have my most current meds list from the other provider.  It was just in an external data view.   The point here is, Innovaccer doesn't have this issue.  We will construct a comprehensive record from all sources and our applications will rely on the most up-to-date data, without all these little hidden points of bias.

NK: You all try to do a lot and offer many products. How do you know when to cut a product? What products unexpectedly succeeded and which ones failed? 

Abhinav: ‍Access Center was a product that took off way faster than expected.

When you call a health system’s phone line to book an appointment, it gets routed to a call center where an agent will pick up, often with very little context on who you are or what your issue is. Call centers are a major revenue driver for large health systems. They’re responsible for keeping each physician’s schedule filled, so maximizing call center productivity is top of mind for every health system. 

Access Center brings together a single unified view of every patient, and our copilot makes it easy for call center agents to find the most relevant information for the ongoing call. It listens to the conversations and surfaces the next best action (e.g., info mined from a knowledge base, or a reminder to book a mammogram due from a previous visit) and then wraps up by summarizing the call, saving the agent a ton of documentation burden. This was a larger problem than we thought (you can see a video on how it works here).

We tend to do better in workflows that require some sort of data-dependency. We rolled out some other applications like communication tools between providers, or virtual care and telemedicine but those haven’t had the same success. Our right to play begins when there is a workflow that has some sort of information asymmetry internally.

NK: How does the 21st Century Cures Act and general push for interoperability impact your business? Which products become more or less valuable?

Abhinav: 

1. Interoperability is a great thing for us.  AI tools, copilots, and insight-enriched workflow support all rely on having good data.  The information blocking aspect of 21st Century Cures, in particular, has helped facilitate access to data within EHR systems. We don't see any of our products being less valuable, with more access to data.  Stated differently, none of our products rely on our ability to hoard patient data, with their value undermined by others being able to access that data.
2. There is uncertainty in the industry right now about the new administration's approach to the TEFCA aspect of 21st Century Cures.  We don't envision any long term major impact here, since there are other national interoperability networks we can rely on where appropriate.  What could have an impact, would be HHS embarking on a new effort, conceptually similar to but different from TEFCA.  Even this, though, would not hurt us per se; it would just have to be evaluated for its impact in the details of our data flow operations.
   

[NK note: For context, at a high-level TEFCA is the government’s blueprint for nation-wide health-information exchange. It’s an attempt to set up rules of the road around what data is available/make it standardized, use cases where healthcare data exchange is allowed, and how the data is actually going to be routed/transmitted. As you can imagine, getting everyone to agree on all of the above is hard + you need entities to actually participate.]

3. Not part of 21st Century Cures, but germane to this topic, are the 2020 and 2024 CMS interoperability rules, which address data flows involving payers.  Their scope is more narrow than that of the Information Blocking Rule, but we are looking forward to what they will enable for prior authorization workflows, in particular (e.g.  requiring payers to support APIs for key elements of the prior auth process)

NK: Innovaccer’s approach seems to be focused on building first party applications on top of the data layer. Is there a world in which third-party apps can build on top of you all today? Or a world where providers “vibe code” their own applications?

Abhinav: We build a lot of our own applications, but we also partner with companies when we don’t have something that customers want.

We think of ourselves more like Apple’s app store, where we curate the vendors we work with.  We already work with 8+ companies we’ve vetted in areas like post-acute care analytics, social determinants data, ADT feeds, etc. We’re also thinking of starting an accelerator by the end of the year for other companies that want to build on top of our foundation + hospital relationships. 

We also see hospitals wanting to build their own very specific workflows in areas where we don’t have as much expertise or it’s not as transferable to other customers (e.g. 340B specific applications). Many organizations look for flexibility to develop tailored tools that reflect their local needs and priorities. We launched a platform called Gravity that essentially provides all the tooling and infrastructure for our customers to build and deploy their own agents and copilots. Whether they need a referral coordinator bot or highly customized ambient scribe, they can do it quickly and without a big development burden. 

There are so many individual agents and point solutions already. EHR interoperability is already hard enough, imagine inter-agentic interoperability. We think we can create a middleware layer that people can develop and exchange information on, in a secure enterprise-grade environment. 

—

Thinkboi out,

Nikhil aka. Nikshnan

Twitter: ​@nikillinit​

IG: ​@outofpockethealth​

Other posts: ​outofpocket.health/posts​

--

{{sub-form}}

‎If you’re enjoying the newsletter, do me a solid and shoot this over to a friend or healthcare slack channel and tell them to sign up. The line between unemployment and founder of a startup is traction and whether your parents believe you have a job.