Published on

Building Healthcare Tech

Opportunities, Challenges & Playbook
Authors

Post Navigation:

Source: Open Source Founder Interview #25 — Cody Ebberson, Co-founder & CTO of Medplum.com

Introduction


Advice for starting in Healthcare Tech

Custom development is a great way to get your foot in the door.

Doctors, administrations & care providers have lots of ideas on improving the healthcare delivery system & patient experience, however they lack a team of software engineers.

Partnering with them to build prototypes, proof of concepts & V1s goes a long way to learn about the space, learn about the opportunity, develop goodwill & get access in the ecosystem.

Watching doctors using your software will unlock countless opportunities to improve the product - it's the most direct & useful feedback.

Being part of the conversation is a necessity - roll up your sleeves, partner up and work in person.

Launching on Hacker News

The advice is get out there and launch. Hacker News is an opinionated audience with lots of thoughts & feelings on how things should be done. Make sure the house is clean & the table is set for a harsh and critical audience to come in and judge your product. Overall the experience was great in terms of both feedback & business development.

Working very closely with customers

Weekly meetings, slack channels, on-site training, guidance on architecture & deployment. Goal is customers love the process & save months of development to open new capabilities & launch faster. Hard but very worth it.

Cloud vs On-prem

15 years ago, cloud used to be a dirty word in healthcare. People didn't believe in healthcare data in the cloud and preferred everything in legacy servers on-prem with physical security. But the model broke down once hospitals started merging, while doctors started wanting to look up patient data on their phones.

The story has now flipped and only cloud is trusted. ex if a device is compromised/lost, the cloud account is killed & the device is rendered useless

There is still a lot of opportunity here to help healthcare organizations transition to cloud-based solutions.

Data Interoperability, Portability & Security

The US government is pushing for healthcare interoperability and data standards, now also offering financial incentives to organizations that support data portability. There is still a lot of opportunity in enabling this. The government is also increasing the required security standards, so there's opportunities to help organizations stay compliant & secure.

Machine Learning & Artificial Intelligence

Don't expect robots to diagnose patients & order tests & meds, that's not realistic in the short term (and that's where a lot of skepticism comes in).

However there are tons of opportunities to automate administrative & operational work at hospitals & doctors' offices (ex answering questions about billing and insurance, rescheduling appointments, updating medical history etc.)

We're in the early days of applying machine learning & large language models in healthcare - there is a lot of opportunity in this space.

Clinical Research, Data Anonymization & Synthetic Data

In medical/clinical research the status quo is tiny data sets, with population sizes of 50-100 patients, due to the lack of access to data or meaningful partnerships.

Potential solutions are data anonymization frameworks & synthetic data.

Anonymization in healthcare is very tricky, namely because it does not take many data elements to uniquely identify an individual (ex job title & ZIP code). Machine learning can help here.

Synthetic data is the process of looking at a real production dataset and using statistical methods & algorithmic processes to create a derivative dataset with the exact same characteristics as the original dataset (ex. gender, age, height, weight, healthcare outcomes, test performed etc.). Statistical guarantees that the synthetic data is representative can lead to population sizes of not 50 but 50,000 patients, which can unlock a lot of value. There's some early players in this space, but there's still plenty of opportunity here.

How to choose customers

Balancing two goals: revenue & key platform capabilities.

Most immediate goal is healthy revenue & a viable business - so opportunities that generate revenue are prioritized.

There's also a strategic goal - work with design partners who help you get access to a new core capability as a platform. Have a list of 10 key tent-post capabilities that the company wants to demonstrate, ex. access to payment networks (receiving payments from Medicare/Medicaid), ordering medications & controlled substances etc. Work with these strategic customers (even for less revenue) because they sponsor the development of these necessary milestones. There are often legal/compliance/bureaucratic challenges that also need to be overcome for every milestone.

Venture playbook: seed round done, decide milestones & metrics for raising next rounds & map what the path looks like, then work backwards.

Healthcare Tech Pricing

Publicly available pricing is a novelty in healthcare tech.

Almost nobody will tell you how much it costs - it's usually 'contact sales', 'schedule a demo' etc. There's a feeling of a sleazy process and meetings after meetings required.

Medplum offers a free tier for its open source product, with public pricing for monthly support and enterprise plans based on support & custom feature development needs of customers.

Silicon Valley Founders, VCs & Venture Studios

There are organizations & VCs specifically focused on helping launch new healthcare companies.

There is also an emerging trend of Healthcare Venture Studios. They provide funding, legal & compliance resources, sometimes marketing & software engineering.

This management tier of healthcare executives with deep experience & connections can help take away some of the risk and guide you during the first years of building your healthcare company.

Be Patient

Building in healthcare can feel slow & over-regulated, but there's a lot of opportunity and demand for people who make the investment to learn how to get things done in this system. Be patient & persist.

Challenging Epic

The most commonly used hospital healthcare platform in the US is Epic.com

The largest hospital system in San Francisco, UCSF (University of California San Francisco), spent $700M on their Epic deployment (public data)

There haven't been enough people challenging this status quo. Surely it will take a new company a decade or two to reach that level, but Medplum.com will (naively) make a run for it!

Medplum.com Founder Pitch

Check out Medplum.com to accelerate your healthcare software development while covering security, compliance & interoperability requirements.

Lots of documentation, example repositories and guides on how to get started & move fast.

⭐ Open Source Code Repository: https://github.com/medplum/medplum

Full Interview & Highlights


Discuss on Twitter