[
Gather AI offers hardware-agnostic machine vision developed at Carnegie Mellon University that can scan barcodes, lot codes, text, expiration dates, and more. The Pittsburgh-based company recently named Joseph Mirabile as vice president of operations.
Mirabile has more than 20 years of experience with robotics, automation, and supply chain innovation.
At Gather AI, he will be responsible for scaling its operations team and overseeing the customer lifecycle, from deployment and support through expansion. The company added that Mirabile will work on implementing its technology stack, which uses physical AI, machine learning, computer vision, and generative AI.
Founded in 2017, Gather AI claimed that its system does not require refits and can work in cold storage, dark warehouses, and very narrow aisles.
Mirabile to apply robotics experience
How does your experience at Seegrid, RightHand Robotics, and Fox Robotics inform your approach to operations at Gather AI? How different are the physical AI tech stacks among these companies?
Gather AI.
Mirabile: My background in warehouse operations gave me a deep understanding of industry challenges. This experience is invaluable at Gather AI, where I address customer issues and deliver solutions. My previous roles focused on autonomous material movement and physical task automation.
Gather AI, however, uses AI-powered drones and MHE [material handling equipment] products integrating cameras and sensors with existing forklifts for comprehensive visibility and real-time intelligence. This allows us to pinpoint misplacements, damages, and expired products.
While past roles offered physical work tools, Gather AI provides the crucial intelligence layer for efficient warehouse management, delivering complete, real-time supply chain visibility, which was a persistent challenge I observed previously.
From your work at Northrop Grumman and FedEx, what are logistics companies looking for now in automation? What pain points still need to be addressed?
Mirabile: Logistics companies are actively pursuing highly intelligent, flexible, and integrated solutions to cultivate resilient, predictive, and hyper-efficient supply chains. They are primarily focused on addressing key challenges such as labor shortages, the escalating demands of e-commerce, and the continuous drive to enhance efficiency and reduce costs.
Gather AI launches MHEV to gather data with forklifts
Aside from recent developments in AI, how mature are the other technologies that Gather AI is working with?
Mirabile: Gather AI initially used drones to validate its technology, successfully proving a significant need for its solution within the confines of every warehouse.
We have since expanded with the launch of our Material Handling Equipment Vision (MHEV) product. This involves installing cameras and sensors onto forklifts that move pallets, allowing us to capture every movement within the warehouse.
This real-time information provides complete visibility into every pallet’s journey, which operations managers can leverage to optimize overall warehouse productivity. Specific benefits include optimizing material flow and providing crucial feedback on damages, inventory counts, expiration dates, and precise locations.
Looking ahead, our technology, including both drones and MHEV, will expand beyond warehouse operations. We plan to utilize this technology on wearable devices and various camera systems to optimize all parts of the supply chain, including cargo ports and yards.
How difficult is it to transform unstructured warehouse data into a single, reliable source of truth for intralogistics? What is Gather AI’s differentiator here?
Mirabile: For decades, the global supply chain has operated with a massive blind spot. While software has digitized nearly every other aspect of business, the physical reality of warehouses — the millions of pallets, cases, and items in constant motion — has remained largely offline.
This gap between what the software thinks is on the shelf and what’s actually there is not a minor inconvenience — it’s a systemic, $1.7 trillion problem that erodes value at every step.
Gather AI is at the forefront of this shift because we uniquely combine three forces:
- Physical AI: We make off-the-shelf hardware — drones, cameras, mobile devices — fully autonomous, turning them into intelligent data-capture agents.
- Deep learning and computer vision: We use state-of-the-art models — running both at the edge and in the cloud — to turn raw visual data into structured, actionable insights with near-perfect accuracy.
- Generative AI: We transform these insights into a strategic, queryable intelligence layer, allowing managers to use natural language to get immediate answers about their physical operations via a dashboard.
Our future-proof software integrates with any setup without infrastructure changes, pairs with off-the-shelf hardware deployable and replaceable within 24 hours, and operates seamlessly across environments — including cold storage, dark warehouses, and narrow aisles — without GPS, Wi-Fi, racking, or retrofits.
Customers expect reliability
How much will your company work with customers on support and feedback? What areas are ripe for more adoption?
Mirabile: Customer support and feedback are central to my role. The Operations team serves as the primary advocate for the customer, and our goal is to gather the insights that drive retention and expansion
The area most ripe for adoption is moving customers from a successful initial implementation to a full, network-wide rollout. My work in gathering feedback is precisely what enables this. By understanding the key features a customer needs to see, we can create the roadmap that justifies their expansion from one facility to their entire network.
What are some of the most common key performance indicators (KPIs) for autonomous systems, and are there some that shouldn’t be overlooked?
Mirabile: When we talk about how well automated warehouses are doing, we usually look at a few key things.
First off, we really care about how reliable everything is and how much uptime we get. That means we’re checking things like how often the system is available, how long it runs before something breaks (MTBF), and how quickly we can fix it when it does (MTTR).
Beyond that, we also keep an eye on how much stuff moves through the warehouse (throughput), how much our customers are actually using our equipment (utilization), how accurate our data is, and how well our team is adopting the new tech.
New VP learning about Gather AI tech
What are some of your incoming priorities for Gather AI?
Mirabile: Operationally, our goal is to scale deployments by establishing an engineering feedback loop. This involves collaborating with the engineering team to share insights on field-discovered bugs, installation hurdles, and future feature requests, all aimed at simplifying deployment processes.
How would you describe the culture, and what are you most eager to learn?
Mirabile: Gather has an awesome culture that really encourages new ideas and taking initiative. Everyone on the team is super collaborative, always ready to jump in and help solve problems or brainstorm fresh ways to make our products and company even better.
I’m especially excited to dive deeper into our tech and explore all the different ways we’ll use it to keep tackling customer issues.
