Humanoid robots have a commercialization problem. Peggy Johnson has to solve it.

Agility Robotics began trading on Nasdaq as AGLT on November 5. The company says it is the first U.S. pure-play humanoid maker with robots already working inside customer buildings. The stock opened above the $10 SPAC reference price, swung from about $8.40 to $12.15, and closed near $11.50. Peggy Johnson, chief executive since March 2024, now has to prove something the ticker cannot invent on its own: that Digit earns its keep after the ribbon-cutting.

She has spent a career turning hard tech into something a customer will renew. After about 24 years at Qualcomm, she ran business development for Microsoft under Satya Nadella, then took Magic Leap out of consumer fantasy and into enterprise sales. Agility, spun from Oregon State University walking-robot research in 2015, hired her in 2024 to do that same translation for Digit, its bipedal warehouse robot. Co-founder Damion Shelton stepped from chief executive to president. Jonathan Hurst, the other co-founder, stayed as chief robot officer. Other labs still chase viral clips of dancing machines. Johnson is selling the shift nobody wants to staff: lift, carry, put the tote on the line, do it again.

At GXO Logistics in Flowery Branch, Georgia, that shift looks concrete. Digit has moved more than 100,000 totes, Agility says, feeding containers onto conveyors beside other automation at a site that has handled SPANX fulfillment. GXO, the world’s largest pure-play contract logistics provider, signed a multi-year robots-as-a-service deal in June 2024 after a late-2023 pilot. The two companies call it the first formal commercial humanoid deployment in live operations. On October 26 they said Digit will also go into GXO warehouses in Statesville, North Carolina, and Lebanon, Tennessee, for e-commerce material handling starting in early 2027. Schaeffler, Toyota Motor Manufacturing Canada, Mercado Libre, and farm-equipment maker AGCO sit elsewhere on the customer list. Across sites, Agility cites more than 65,000 hours of real-world work and nine committed facilities. The robot has a job description. It still has a thin receipt book.

The invoice is easier to read than the income statement. In Agility’s SEC illustrations, a customer pays about $25,000 to put each robot on site, then about $8,500 a month for access, software, and maintenance. Buying the hardware instead costs about $200,000 plus deployment, with roughly $36,000 a year still due for software and support. Hurst has said the monthly rate is meant to sit near what a warehouse already pays people, so the math can work if the robot keeps pace on a long shift. Those figures are Agility’s pitch, not a customer’s audited return. A logistics buyer can start writing checks. Agility still has to turn those checks into a company that covers its own burn.

That gap shows up fast in the prospectus. Net sales were about $1.8 million in 2025 against a $138 million loss, and related parties made up a large share of sales. In the first half of 2026, sales were about $174,000 and the loss about $362 million, helped along by stock-based compensation. Digit can collect rent at GXO and still look almost pre-revenue from outside. The $300 million-plus in multi-year Digit 5 orders Johnson points to are potential value over time, and only if milestones, features, and specs get hit. They are not this quarter’s revenue. Digit 4 mostly works inside fenced cells. Digit 5 is built to work nearer people, with early customer access aimed for the first half of 2027 and broader availability by year-end. Uncaging the robot is how Agility hopes to sell more than a few machines per building.

A camera-ready tote pick is the easy part. A few weeks with engineers on site is still a pilot. A real install has to stay up, hit accuracy targets, keep pace with the conveyor, and survive the first renewal meeting after the novelty wears off. Agility says Digit has posted about 98% on-task accuracy at key sites. Hurst has called the jump from “good enough for video” to factory-grade reliability an order of magnitude harder. GXO’s leadership has said humanoid returns are still years away even as it adds Digit sites. Johnson’s Customer Acceleration Program is the slow path she wants customers to walk before Digit 5 ships: prove the workflow, then commit the floor space.

The public listing buys her runway and a scoreboard. The merger with Churchill Capital Corp XI and related financings brought about $465 million in gross capital before expenses, including a Foxconn-led private investment, with Nvidia and Amazon among the wider backers. The deal valued Agility around $2.5 billion before the close. RoboFab, the company’s Salem, Oregon factory, is built for up to 10,000 robots a year if production ever fills it. Near-term company materials still talk about hundreds of Digit 5 units in 2027 if the milestones hold. Public investors can stack that against Figure, Tesla’s Optimus calendar, and Boston Dynamics’ Atlas plans. Johnson’s claim is quieter. Named warehouses. Hours on the clock. A monthly fee.

If someone in your family already works a dock or a plant floor, Digit is not aiming at every job in the building. It is aimed at the tote-and-parts loops that are hard to keep filled. If AGLT shows up in a brokerage app next to other physical-AI names, size it like any early public robotics story with Taxable Investing Basics for First-Gen Professionals, and keep a separate vest calendar with RSU Vest Day Playbook for Diaspora Households. For the other big humanoid leadership move this fall, see Rohit Prasad takes Boston Dynamics. Hyundai wants room for 30,000 Atlas robots a year..

From here, watch the ordinary stuff. Whether the new GXO sites stick, whether Digit 5 clears its gates, and whether Agility’s sales climb off six-figure halves toward the order book on the slide deck. Johnson does not need another robot that trends for a day. She needs customers who keep paying after the cameras leave.