Gather AI Raises $40M for Warehouse Drone Technology
▼ Summary
– Gather AI, a warehouse automation startup, has raised a $40 million Series B funding round led by Smith Point Capital, a firm founded by former Salesforce co-CEO Keith Block.
– The company uses off-the-shelf cameras and drones equipped with a unique “curious” AI to monitor warehouse operations and log data into management systems.
– Its AI is based on classical Bayesian techniques combined with neural networks, not large language models, allowing it to learn from visual data without hallucinations.
– The technology scans for specific items like barcodes and expiration dates to discover issues such as low inventory or misplaced stock, even in harsh environments like freezers.
– The startup, founded by Carnegie Mellon PhD alumni, is positioned in the “embodied AI” field and has now raised $74 million total, serving customers like GEODIS and NFI Industries.
A significant $40 million Series B investment round has propelled Gather AI, a company specializing in autonomous warehouse inventory drones, into its next phase of growth. The funding was spearheaded by Smith Point Capital, the venture firm established by former Salesforce co-CEO Keith Block. This capital infusion brings the startup’s total funding to $74 million, signaling strong investor confidence in its unique approach to solving warehouse logistics challenges. The company’s technology aims to enhance accuracy and safety by providing real-time visibility into inventory and operations.
The connection with lead investor Smith Point Capital was forged at a logistics conference roughly a year ago. According to Gather AI co-founder and CEO Sankalp Arora, the firm’s team grasped the startup’s vision almost immediately. “It took Keith and his team five minutes to get what we’re doing,” Arora noted. The founders’ background is deeply rooted in advanced robotics, having first collaborated as PhD students at Carnegie Mellon University. There, they developed one of the earliest autonomous helicopters, conducting tests on the FBI training grounds in Quantico. Keith Block serves as a trustee for CMU, adding a layer of connection to the academic institution where the core technology was born.
Launched in 2017, Gather AI translates the founders’ expertise in autonomous flight into practical warehouse solutions. The system utilizes readily available cameras mounted on equipment like forklifts, alongside commercial drones that navigate warehouse aisles. These devices continuously monitor floor activities, capturing data that is seamlessly integrated into existing warehouse management systems. The innovation lies not in random data collection, but in what Arora describes as a “curious” artificial intelligence. This curiosity drives the system to actively seek out and interpret specific visual information, from barcodes to text and physical conditions.
The AI’s focus extends beyond simple scanning. It is programmed to identify lot codes, expiration dates, case counts, product damage, storage occupancy, and other critical markers. This enables the platform to detect and even predict operational issues such as low inventory levels, misplaced stock, and workflow patterns that could lead to safety concerns. The technology is particularly valuable in environments that are challenging for human workers, such as industrial freezers and cold storage facilities, where it can operate reliably without interruption.
It is important to distinguish Gather AI’s technology from the large language models that dominate current headlines. The system is not built on end-to-end neural networks. “They are classical Bayesian techniques, combined with neural networks,” Arora explains. These probability-based methods teach machines to interpret visual data by leveraging prior knowledge and new information, a process that avoids the “hallucination” problems sometimes associated with LLMs. This foundational approach allows the AI to learn from its environment, gather pertinent information, inspiring the company’s name, and decide on subsequent actions based on its findings.
This methodology positions the company at the forefront of what is often termed “embodied AI,” where intelligent systems interact directly with the physical world. This contrasts with language models that operate purely in digital spaces. The startup’s progress in this field was recognized with the 2025 Nebius Robotics award for Vision AI and Streaming Video Analytics. Currently, Gather AI serves a roster of enterprise clients including Kwik Trip, Axon, GEODIS, and NFI Industries, and employs a team of approximately 60 people. Alongside Smith Point Capital, other investors in this round include Bain Capital Ventures, XRC Ventures, and Hillman Investments.
(Source: TechCrunch)
