Ambi Robotics
Integrator / OEM
Parcel sortation and induction robots (AmbiSort, AmbiStack) running on the Dex-Net-derived AmbiOS sim-to-real stack.
Ambi Robotics builds AI-driven parcel sortation and induction robots for e-commerce, postal and third-party logistics operations. Its systems — AmbiSort for small-parcel sorting and induction, and AmbiStack for mixed-case stacking — run on AmbiOS, the company's simulation-to-reality operating system descended from UC Berkeley's Dex-Net research. The pitch is dexterity: robots that pick items they have never seen without retraining. Ambi reports its systems deployed in the majority of U.S. states, across major retailers, parcel carriers and Fortune 500 logistics companies, with more than 150 million objects sorted over five years and roughly $70 million of venture capital raised to date.
Research mapping
Physical AI / Sim-to-Real, Digital Twins & Validation (L9) / Integrator / OEM
Evidence boundary
Original analysis and independent corroboration remain under development. Commercial scale and durable technical advantage remain under review.
What makes it defensible
Dex-Net-derived sim-to-real grasping plus years of parcel-level manipulation data lets Ambi pick unseen items without retraining — the dataset, not the arm, is the asset.
Position in the stack: Integrator at the warehouse cell: it assembles robots, grippers and the AmbiOS learning stack and sells throughput to parcel and 3PL operators, owning the deployment data that improves each policy.
Who buys it
Founding team
Jeff Mahler — Co-founder & CTO — Created the Dex-Net project at UC Berkeley; PhD in EECS (2018); previously co-founded Lynx Laboratories, acquired by Occipital.
Stephen McKinley — Co-founder & VP Operations — PhD in mechanical engineering, UC Berkeley (2016); postdoctoral fellow with Pieter Abbeel.
David Gealy — Co-founder & VP Hardware — Robotics researcher at UC Berkeley in the Robot Learning Lab and AUTOLAB; MS/BS mechanical engineering.
Matt Matl — Co-founder — UC Berkeley doctoral graduate; co-founder of Ambidextrous Laboratories, later Ambi Robotics.
Ken Goldberg — Co-founder & Chief Scientist — Professor at UC Berkeley (William S. Floyd Jr. Distinguished Chair in Engineering); PhD in computer science, Carnegie Mellon; led the AUTOLab where Dex-Net originated.
SourceFinancing detail
- Last round
- Series B (2022)
Investors: Tiger Global, Bow Capital, Vertex Ventures US
Mapped funds and investors
Direct relationships supported by published evidence.
Reported holdings as-of
No sourced relationships mapped yet.
Disclosed investment history
No sourced relationships mapped yet.
Showing up to four relationships in each group. ETF weights describe the fund’s portfolio, not ownership of this company. Private investment history does not establish current ownership.