Our take
OSARO is one of the more proven AI picking vendors for e-commerce bagging. Its SightWorks perception software uses deep learning models trained on millions of items, so robots handle new SKUs without manual registration.1,5
The best evidence is in production: a bagging system at Korea's largest e-commerce company, sized for more than 1,500 autobaggers, and a FANUC-built cell at Zenni Optical that raised throughput 80%. The weak points are quote-only pricing and little public information on revenue or recent funding.6,8
Scores by criterion
Each criterion is defined on our methodology page.
Who uses OSARO and what it's for
E-commerce retailers, fulfillment centers, and their integrators use OSARO for robotic bagging, piece picking from AMR and ASRS systems, kitting, and mixed-case depalletizing.1,3
Named users include Coupang in Korea and eyewear seller Zenni Optical in California.6,8
History
Derik Pridmore, a former Founders Fund investor, and Michael Kahane co-founded OSARO, which says it has worked on AI-driven robotics for more than 10 years. Its early research covered deep reinforcement learning and imitation learning from video demonstrations.2,1
In March 2025 Fast Company named OSARO to its list of the World's Most Innovative Companies, in the top 10 for robotics and engineering. In August 2026 it completed a SOC 2 Type II audit.4,3
SightWorks and robot cells
SightWorks generates confidence maps from RGB and 3D data to find objects and grasp points, and AutoModel learns new items on the fly. Pridmore says newer models trained on production data need less learning at each new install.5,1
The Robotic Bagging System pairs OSARO picking with automatic bagging machines from established suppliers, and the HyperCare program monitors systems after go-live.4,1
Pricing
OSARO does not publish prices. Systems are scoped per site, and the company invites buyers to send products for a test or visit a live deployment.2
Competitive landscape
| Company | Piece-picking product | Pricing model | Deployment evidence | Where OSARO wins or loses |
|---|---|---|---|---|
| OSARO | SightWorks perception and robot cells for bagging and piece picking from AMR and ASRS systems1 | Not published; scoped per site2 | Coupang bagging program sized for 1,500+ autobaggers6,8 | Bagging and piece picking at e-commerce scale; prices not published |
| RightHand Robotics | RightPick autonomous piece-picking cell, cycle times as fast as about 3 seconds9 | Not published; system quotes9 | Rockwell Automation strategic investment in 202510 | RightHand wins on focused RightPick cells with fleet software and a Rockwell channel. OSARO wins on e-commerce bagging at large scale. |
| Nomagic | Pick, Pack, and Sort stations at AutoStore, shuttle, and pocket-sorter ports; up to 500 picks per hour12 | RaaS monthly fee or CAPEX with SaaS; rates not published13 | Zalando rolling out up to 50 robots after a 100,000 picks per day pilot11,22 | Nomagic wins on storage-agnostic stations with RaaS terms and large fashion rollouts. OSARO wins on e-commerce bagging at large scale. |
| Mujin | MujinOS-controlled piece picker, up to 1,000 pieces per hour14 | Not published; inquiry and quote23 | More than 20 robots in one center; ASKUL, LOGISTEED, and ELECOM named14 | Mujin wins on a rated 1,000 pieces per hour and multi-robot centers in Japan. OSARO wins on e-commerce bagging at large scale. |
| KNAPP Pick-it-Easy Robot | Picking station fed by KNAPP shuttle storage, items up to 3.5 kg16 | Not published; part of a KNAPP project quote16 | EDEKA NORDfrische runs two robots at up to 450 picks per hour each17,16 | KNAPP PiE wins on native fit inside KNAPP storage and software, with group backing. OSARO wins on e-commerce bagging at large scale. |
| Swisslog ItemPiQ | KUKA-arm station with native AutoStore and SynQ integration, typically up to 600 items per hour20 | Not published; part of a Swisslog project quote19 | No named current customer on the product pages19 | ItemPiQ wins on published specifications and native AutoStore and SynQ integration. OSARO wins on e-commerce bagging at large scale. |
Limits
Pricing is quote only.2
Revenue and recent funding are not disclosed. We estimate revenue at $10 million to $25 million, based on production deployments in five countries and a Coupang program sized for more than 1,500 autobaggers. This is an estimate, not a reported figure.1,6
OSARO supplies the intelligence and works through robot makers and integrators such as FANUC, so buyers often deal with more than one vendor.8
What customers say
Where it helps
At Zenni Optical, robot cells pick each pair of glasses in its case, scan the barcode through the case, and drop it into a bagging machine. Zenni reports an 80% increase in throughput.8
Coupang's fulfillment centers run 18 to 20 hours a day with more than 250,000 SKUs, and OSARO set up a five-person engineering team in Korea within four months to support them.6
Where it falls short
Parcel hubs that mainly need high-speed sortation of boxes and polybags may get more from purpose-built parcel systems such as Ambi Robotics or Plus One Robotics.
Leadership

Derik Pridmore
CEO and co-founderPridmore studied physics and computer science at MIT and invested at Founders Fund, where he backed DeepMind, before founding OSARO.2,7

Michael Kahane
CTO and co-founderKahane was a senior engineer and applications manager at Samsung semiconductors.2

Sources
- OSARO, Homepage, accessed November 3, 2026
- OSARO, About us, accessed November 3, 2026
- OSARO, News and media, accessed November 3, 2026
- OSARO, OSARO named to Fast Company list of the World's Most Innovative Companies of 2025, March 21, 2025, accessed November 3, 2026
- Robotics 24/7, OSARO SightWorks advances AI-powered robotic picking and packing, October 13, 2026, accessed November 3, 2026
- RoboticsTomorrow, OSARO's AI-driven automation enhances fulfillment speed and accuracy, February 5, 2026, accessed November 3, 2026
- MIT Technology Review en español, Así es el creador de uno de los robots más inteligentes del mundo, accessed November 3, 2026
- FANUC America, How Zenni Optical built a faster, smarter, more scalable fulfillment line, accessed November 3, 2026
- RightHand Robotics, Homepage, accessed November 6, 2026
- RightHand Robotics, Rockwell investment, accessed November 6, 2026
- Zalando, Zalando to install up to 50 AI-powered Nomagic robots, March 19, 2026, accessed November 9, 2026
- Nomagic, Pick solution and FAQ, accessed November 9, 2026
- Nomagic, Partners page, RaaS and CAPEX with SaaS models, accessed November 9, 2026
- Mujin Japan, MujinRobot piece picker product page (Japanese), accessed November 9, 2026
- Mujin, Mujin raises US$233 million, December 2, 2025, accessed November 9, 2026
- KNAPP, Pick-it-Easy Robot product page, accessed November 9, 2026
- KNAPP, Pick-it-Easy Robot at EDEKA NORDfrische case study, accessed November 9, 2026
- KNAPP, KNAPP continues its growth trajectory, fiscal 2025/26 results, accessed November 9, 2026
- Swisslog, ItemPiQ robotic piece picking product page, accessed November 9, 2026
- Swisslog, ItemPiQ factsheet (PDF), accessed November 9, 2026
- KUKA, Annual Report 2025 (PDF), Swisslog segment and group sales, accessed November 9, 2026
- Nomagic, ASOS case study, accessed November 9, 2026
- Mujin, Piece Picking application page, accessed November 9, 2026


