Transforming Healthcare Inventory Management: How Z5 Inventory Harnessed AI and AWS to Revolutionize Our Counting Process
At Z5 Inventory, we believe hospitals should never waste time or money on outdated, inefficient inventory practices. Since our founding in 2015, we’ve been obsessed with helping healthcare organizations reduce costs, cut waste, and get the right supplies to the right place at the right time.
Every year, U.S. healthcare facilities lose billions of dollars due to expired, excess, or misplaced medical supplies. This isn’t just a financial problem — it impacts patient care and operational readiness. Our goal has always been to eliminate that waste and empower hospital teams to focus on patient outcomes rather than supply headaches.
Even though we’d already moved far beyond pen-and-paper spreadsheets, our previous system still had major bottlenecks:
Video-based Counting: We armed our teams with Go Pros and mobile apps to quickly record inventories of medical and surgical supplies.
Manual Transcription: After recording, staff had to painstakingly watch and transcribe every item mentioned in 45–60 minutes of footage per inventory session.
High Costs and Errors: This process was slow, expensive, and prone to mistakes. More hospitals meant more staff hours and higher costs.
We knew there had to be a smarter way. Our goal was to automate transcription and data extraction so our platform could generate reliable, actionable inventory records without the human bottleneck.
We decided to work with Caylent, a cloud and AI solutions provider, to help us design a cutting-edge system powered by AWS technologies. Our objectives were clear:
Automation: Eliminate the manual transcription step entirely.
Speed: Reduce turnaround time from hours to minutes.
Accuracy: Deliver highly reliable data that improves decision-making.
Security: Stay fully HIPAA-aware and protect sensitive information.
Scalability: Handle growth without adding staff headcount.
Together with Caylent, we built a next-generation inventory processing pipeline. Here’s how it works:
Video Capture: Our teams record inventory counts just as before.
Automated Transcription with Amazon Transcribe: Speech from the videos is instantly converted into text.
Structured Data with Amazon Bedrock & Claude 3.7 Sonnet: The transcription is transformed into standardized, machine-readable inventory data, ready for our backend systems.
Scalable Infrastructure: Using AWS Lambda, Step Functions, SQS, and S3, we process large volumes of video concurrently — no more queuing or manual delays.
Secure Handling of Data: All data is encrypted; PHI is excluded from the process; workflows are designed to stay compliant with healthcare regulations.
Seamless Integration into Our Platform: The structured data automatically appears in our dashboards, giving our team and customers instant visibility.
By replacing manual labor with automated AI pipelines, we’ve dramatically shortened the time between recording inventory and seeing actionable data.
Since launching our new system, we’ve seen dramatic improvements in every area we care about:
5× Faster Turnaround: From 45–60 minutes per video down to about 1–2 minutes.
92% Lower Transcription Costs: Freed up budget to invest elsewhere.
Accuracy Boost: Inventory accuracy trending toward 95–98% (up from ~85%).
Massive Scalability: We can now handle more hospitals, more locations, and more videos without adding headcount.
Better Customer Experience: Hospitals get actionable inventory insights almost immediately, enabling better decisions about ordering, stocking, and rotation.
This transformation has changed the way our own team operates and — most importantly — the way our healthcare customers manage their supplies.
Healthcare organizations trust Z5 Inventory to simplify a task that’s historically been painful and time-consuming. By leveraging AI and AWS, we’re now able to:
Cut Wasted Time: Hospital staff spend less time counting and reconciling supplies.
Reduce Waste: Faster, more accurate data leads to fewer expired or overstocked items.
Improve Cash Flow: Hospitals can order smarter and free up capital tied in excess inventory.
Focus on Care: Staff can shift their attention back to patient care rather than supply chain tasks.
This project marks the beginning, not the end, of our innovation roadmap. We’re already exploring:
Pharmacy Inventory Management: Applying our approach to pharmacy stock, where standardized identifiers and packaging can deliver even greater accuracy and speed.
Audio-Only Transcription: Allowing for faster processing and fewer compliance concerns by focusing on verbal counts rather than full video.
Predictive Analytics: Using our growing dataset to help hospitals forecast needs and optimize inventory levels proactively.
Smarter Dashboards and Insights: Giving healthcare leaders richer, real-time data to make strategic decisions about their supply chain.
At Z5 Inventory, our commitment to our customers is simple: We will keep pushing to make healthcare inventory management faster, smarter, and more cost-effective. By embracing AI and the cloud, we’re proving that even long-standing industry challenges can be transformed with the right technology and the right partners.