In just two years, Tony Zhao and Cheng Chi have built Sunday Robotics, a $1.15B startup using real human data to develop a home robot named Memo.
Tony Zhao and Cheng Chi are betting that the next big robotics breakout will be a robot that can survive a messy kitchen. Their startup, Sunday, is building a home robot called Memo to take on the chores people don’t like doing – and investors are buying in.
Just two years in, the Mountain View-based startup has reached a $1.15 billion valuation. It came following a $165 million raise in a recent funding round led by Coatue Management.
Behind that milestone is a shared foundation that took shape in academia. Zhao traces his roots to Stanford’s computer science PhD program, where he worked on robotics projects like ALOHA and ACT before leaving to build Sunday.
Chi, Sunday’s CTO, took a similar path. He completed a PhD in computer science at Columbia before continuing his work at Stanford, where he focused on problems related to controlling physical systems through software.
Memo, the household robot designed for everyday chores
Their shared instinct is that the hardest part of robotics is programming reliable behavior. Both founders have been linked to work that helped popularize teaching robots through human demonstrations.
With Sunday, Zhao and Chi have taken that approach and applied it to a familiar frustration with chores. Their flagship humanoid robot, Memo, is designed to handle everyday tasks like loading a dishwasher or helping with laundry.
It may sound straightforward, but homes can be unpredictable. Layouts differ, appliances change, counters get cluttered, and routines rarely look the same from one day to the next.
That’s why Sunday has been unusually open about what it thinks matters most: training on behavior that looks like real life.
How Sunday is training robots using real human behavior
One of its signature methods is glove-based data capture. Last year, WIRED reported that Sunday was paying people to do chores while wearing gloves designed to resemble Memo’s hands. Those workers, also known as “Memory Developers”, then record detailed movements to train the system. Zhao told WIRED the gloves cost about $400 per pair.
Sunday calls this a kind of feedback loop: the more real-world behavior it collects, the better its system becomes. It echoes their goal to move beyond the staged demos that have long defined home robotics and toward something that holds up.
It is part of a broader push to bring robots into everyday homes, a shift that some in the industry see as a turning point for consumer robotics.
Other startups are exploring similar ideas, from humanoid assistants to systems that learn tasks by watching humans. 1X Technologies, an OpenAI-backed startup, is building a $20,000 robot called Neo that learns by watching videos.
For Sunday, this kind of funding round now shifts the focus from promise to proof. The startup has only recently emerged from stealth. And while it has a reported waitlist of around 1,000 people, attention alone doesn’t always validate an approach.
