For those who make America.

We're building foundation models for physical AGI by automating retail fulfillment with American-supervised robots in American stores.

Partner with us Our robotics

Robotics

Bridging the gap between AI and your business.

Human-supervised robots pack online grocery orders in your existing stores.

The stakes

Online order economics are broken

Picking labor is the problem: every hand-packed order carries $8-12 of labor on a business that nets ~2.1%.

$8-12 Labor per hand-packed online order, roughly 10% of the ticket.
~48% Annual picker turnover. Wages are up ~20% in five years.
75% cheaper Revere's robotics solutions pack online orders under $2.8 all-in, in your existing footprint.
“Manual store-based fulfilment tends to hit a ceiling at around 12-15% online penetration.” McKinsey & Company, The State of Grocery North America 2026

The trend

Online became one dollar in five of grocery

+20% vs <2% Online vs. store growth across six straight quarters.
~75% Of 2025 grocery dollar growth was e-commerce.
27-37 min Staff time per online order, up to 125% more labor per basket.

The offer

More orders per labor hour

Our economics let you expand the market for digital orders without CapEx, and with the right to cancel any month.

Typical metro staff

~60¢

per item

~$12 on a 20-item order
$25/hr · ~60 items/hr

Revere robotics

20¢

per item

$4 on a 20-item order
Pay per order · no CapEx

  • Reuse the store you already have No dark-store buildout. 30,000-140,000 SKUs stay where they are.
  • American-supervised robots Remote U.S. operators supervise every pick while autonomy compounds.
  • Ride the automation curve down Start on the night shift tonight. Cost falls as supervision retires.
  • Price follows hours of operation As low as 6¢/item day+night; 25¢ nighttime only. We amortize the robot.
Talk robotics

How we get there

Useful work today. Autonomy next.

We deploy supervised robots into messy retail environments to solve fulfillment now, and train foundation models for physical AGI on the contact of real work.

  1. Deploy on real jobs

    L2 robots in stores, supervised remotely by American operators.

  2. Solve fulfillment

    Pack online orders cheaply tonight. Value before solo autonomy.

  3. Train on contact

    Every hour becomes force, tactile, proprioception, and social context.

  4. Climb L2 to L4

    Retire supervision step by step as capability compounds.

Retail is the most diverse environment for true scaling: contact and social diversity that labs and household stacks cannot buy.

The problem

Retail fulfillment is still human work

Stores are unstructured, occupied, and unforgiving. Stocking, picking, and restocking demand contact with the world: force, touch, judgment in crowded aisles that today's robots cannot yet do alone. Labor is scarce. The work does not wait.

Our hypothesis: the path to physical AGI runs through automating retail fulfillment, training foundation models on the contact of real American work.

Partner with us

The Revere path

Work first. Autonomy next.

  1. Deploy on real jobs

    Robots in stores, supervised and operated by remote American employees.

  2. Solve fulfillment

    Deliver the business outcome (stocking, picking, restocking) before models are ready to fly solo.

  3. Train on contact

    Every hour of work becomes force, tactile, proprioception, and social context for our foundation models.

  4. Retire supervision

    Step by step, capability compounds until human oversight can step back.

We work the hardest environments, unstructured, occupied, unforgiving, because that's where physical AGI has to learn.