PearX recently hosted its latest demo day in San Francisco, featuring a fresh lineup of startups that caught the attention of venture capitalists. PearX, a 12-week program run by Pear VC, a pre-seed and seed-focused venture firm, limits its cohorts to just 20 startups. This selective approach cultivates a high-caliber environment. PearX’s bi-annual demo day is consistently attended by top venture capitalists, reflecting the historically strong caliber of companies in each batch. During the demo day, 16 startups presented, and five made a notable impression on investors.
Speridlabs wowed attendees with its spatial foundational models designed for robotics, gaming, and special effects. While it’s not the only contender in this field, Speridlabs sets itself apart by claiming that its competitors–such as Runway, Odyssey, and Google’s Genie–lack the ability to be queried or modified. The company’s solution, Mundus, is likened to a 3D Midjourney, enabling persistent geometry even when components are altered.
Saia drew attention with its groundbreaking chip design that performs inference directly on devices. As traditional GPUs and TPUs from Nvidia and Google struggle with high memory costs and power consumption, Saia claims its chip can function using flash storage, achieving significantly faster speeds and eight times the capacity while consuming four times less power than Nvidia’s Jetson. Saia plans to start fabricating test chips next year and aims to launch mass production by 2028.
Ren is dedicated to developing a more secure AI personal assistant. It claims to provide functionalities similar to Muse and Instinct but places a stronger emphasis on security and privacy. Ren keeps data on-device whenever feasible or uses a secure private cloud, ensuring that user actions comply with strict guardrails set by users. Unlike competitors that rely on human operators for tasks like making phone calls, Ren utilizes a fully automated voice system to protect user privacy.
Veros seeks to simplify trust and estate planning, a typically complex and costly process. By harnessing AI, Veros aims to automate tasks usually handled by attorneys and trust administrators. The startup is already managing $250 million in AUM and is in the process of securing a trust charter to operate as a regulated trust company.
Datum, another startup, is focused on accelerating product development in industrial design. The company employs its proprietary Geometric Fingerprint technology to index a firm’s library of 3D designs. This technology assists engineers in identifying parts based on shape, streamlining and automating the engineering process across a broad market in physical design.
