Hi everyone, we are Abu and Tom, co-founders of Astrix Health

TL;DR

Astrix builds AI purchasing agents for healthcare providers.

Our agents automatically purchase the right products at the right time, helping eliminate the $25 billion in wasted spend on medical supplies annually in the US.

Why is this important?

The US spends roughly $4.5 trillion per year on healthcare, with wasteful spending accounting for up to a quarter of that amount. On average, hospitals in the US waste at least $12 million each on medical supplies. Adding to the problem, nurses can spend anywhere between 20 minutes to one hour per shift searching for medical supplies.

For clinics, this problem is even worse. Often, the same people responsible for treating patients are also in charge of purchasing supplies. They’re already stretched on time to care for patients, let alone shop around for the best deals.

While shadowing an imaging center, we observed a nurse doubling as an office manager, simultaneously managing inventory, searching for deals, placing orders, and treating patients. When we got home we started building.

Our solution

Our product has three main functions:

1. Analyze: The agents comb through purchasing data (such as invoices) to identify high-spend areas and cost drivers.
2. Search: Deployed AI agents search for better deals, considering provider-specific arrangements like preferred vendor discounts, GPO contracts, rebates, and other incentives.
3. Purchase: Where appropriate, Astrix place orders when more cost-effective options are found, saving providers time and money.

In under three weeks, Astrix has started working with 10 healthcare clinics, identifying an average 20% cost savings across our first users’ recurring supply purchases. Our AI agents deliver immediate value, reducing expenses and freeing up providers to focus on patient care.

Why now?

Until recently, it was impossible to efficiently scan numerous websites and compare the myriad factors influencing purchasing decisions—such as per-unit costs, availability, expiration dates, device sizes, shipping, and more—at scale to find the best deals. Automating ERP and website navigation for purchases or negotiating with vendors was also previously unfeasible.

Additionally, we're frustrated by the poor quality and outdated nature of the incumbent supply chain software and strongly believe providers deserve better. AI offers a unique opportunity to fundamentally rethink how healthcare professionals interact with computer systems.

Moving forward

Purchasing is just the beginning. We aim to become the comprehensive optimization layer for hospital supply chains, integrating inventory, purchasing, and patient data to put procurement operations on autopilot—ultimately reducing that $25 billion waste to zero.

Healthcare professionals should never feel stressed about ordering or obtaining medical supplies, nor should patient care be delayed due to supply chain inefficiencies. That’s the problem we’re committed to solving.

The team

Our team has extensive experience building AI/ML systems for major healthcare institutions. Abu and Tom, who met at Palantir, developed some of the first generative AI projects for the UK's National Health Service.

Abu led Palantir's medical and pharmaceutical distribution vertical in the US and its healthcare generative AI team in the UK. Previously, he managed data projects at Duke Health and MedStar Health across clinical science projects and hospital operations.

Tom has experience solving optimization problems with AI from his time as an Engineer at Palantir and his Maths degree from Oxford.

Our ask

Email us at founders@astrix.health. We're seeking introductions to:

• Hospitals
• Private practices
• Ambulatory surgical centers

We'd be incredibly grateful for any connections. As a bonus, if they share their last month's purchasing data with us, we'll provide a detailed analysis of their potential cost and time savings.