Jon Goldsmith and the Trouble With Dinner

Dinner has an awkward habit of requiring agreement. Someone wants a burger. Someone else wants something entirely different. The person collecting the orders would quite like everyone to decide before tomorrow. Jon Goldsmith built a company around the possibility that this small domestic negotiation deserved a better arrangement.

Local Kitchens lets people combine dishes from different restaurant brands in one order, prepared by one kitchen team. The ambition sounds modest when described from the dining table. From behind the counter, it involves recipes, staffing, training, ordering software, and the minor miracle of having everything ready together. A hamburger has no interest in waiting politely for another cuisine.

Goldsmith came to this problem through software. He had worked at DoorDash in high school, co-founded an artificial intelligence company, and moved through work in robotics and legal technology before starting Local Kitchens with Andrew Munday and Jordan Bramble in 2020. His career offers an unusual route into hospitality: a succession of attempts to make complicated human activity easier to coordinate.

Before the kitchen, a telephone call

Goldsmith was born in San Francisco and says he still lives there. In his own account of his early career, he was DoorDash’s first software engineer while still in high school. It is a striking beginning for someone who would eventually become responsible for actual kitchens. He encountered restaurant technology from the side concerned with getting an order from one place to another.

At Phillips Exeter Academy, Goldsmith and Nikolas Liolios began experimenting with machine learning. The hobby became Tetra, a service that turned phone conversations into searchable transcripts. Its practical appeal was easy to understand: people could concentrate on a conversation without also trying to preserve every useful sentence in their notes.

Tetra joined Y Combinator’s Winter 2017 batch. By the time it appeared on the 2018 Forbes 30 Under 30 Enterprise Technology list, it had raised $1.5 million. Asset managers, journalists, and recruiters were among the customers described at the time. The founders were already thinking about extending the product to video calls and meetings held in person.

Goldsmith says he skipped MIT to start and sell an AI company; Y Combinator lists Tetra as acquired. The choice belongs in his story because it places building before credentials. He had a working problem and a company to pursue. The education could wait. His subsequent work would keep moving between software and the arrangements people depend on to get things done.

The problem hiding behind a restaurant’s second address

At DoorDash, Goldsmith worked in engineering and Munday in operations. They later reunited as co-founders. Goldsmith met Bramble while working as a product manager at Atrium, bringing another earlier connection into the business that became Local Kitchens. The founding team assembled around a mixture of technical and operational experience.

The restaurant owners they spoke with had a recurring complaint about expansion. Opening another location consumed months of their lives. A spare kitchen address would solve only part of that problem. Someone still had to hire, manage, buy ingredients, and keep the operation running. The extra restaurant could demand the owner’s presence just when the original one needed it too.

Local Kitchens took responsibility for that work. It hired its own employees and operated the locations, with partner restaurants training the teams to reproduce their food. The partners supplied recipes and brands; Local Kitchens supplied an operating business around them. That division of labor gave the venture a more demanding job than simply providing space.

The suburban setting mattered. A restaurant might have devoted customers who had moved farther from its original neighborhood. Goldsmith and his colleagues wanted those customers to find familiar food nearer home. The proposition also appealed to a family whose members shared an address and very little enthusiasm for one another’s dinner choices.

By April 2021, the company had locations in Lafayette, San Jose, and Cupertino. Early partners included Curry Up Now, Proposition Chicken, Señor Sisig, and Wise Sons. Menus were shortened to selected dishes. Ingredients, recipes, and cooking methods had to remain recognizable to customers who already knew the originals. Convenience would have limited charm if the familiar sandwich tasted unfamiliar.

“Our perspective is there’s already tons of great food out there.”

Jon Goldsmith, 2022

A dining room is also an engineering problem

The physical spaces put limits on the software ambitions. William Duff Architects helped develop the first Lafayette location, and later worked on the Palo Alto restaurant. Each kitchen needed a layout and workflow that could support several brands. At Palo Alto, the existing kitchen was expanded while the public-facing space received updated branding and design.

This is where Goldsmith’s career becomes more interesting than a simple progression from computers to food. The work kept crossing the boundary between the two. A screen can accept a complicated order instantly. A cook must still make it. The layout of a station, the length of a menu, and the timing of an instruction all become part of the same customer experience.

Local Kitchens built much of its own technology, including its point-of-sale and kitchen display systems. One program, called Dynamic Firing, coordinated preparation so dishes from different brands could finish together. That is a distinctly practical use of software. Nobody at the table needs to admire the scheduling system. They would prefer their dinner to arrive.

Hospitality had a place in the model from its early years. Locations employed greeters to help guests order and make recommendations. Some offered indoor or outdoor seating. Even a business built to simplify pickup needed someone who could answer a question without directing the customer to another screen.

The money had to reach the stove

In June 2021, Goldsmith announced a $25 million Series A led by General Catalyst, with participation from investors including Human Capital, Pear VC, Fifth Wall, and Penny Jar. He wrote that the company had helped more than a dozen local restaurants generate millions in additional sales in less than six months since its seed round. These were company-reported results, and they explained what he wanted the business to accomplish.

His argument was that local chefs should have access to a growth model that worked economically, rather than leaving expansion to chains with larger resources. Penny Jar’s investment account made a related point: familiar restaurant brands could bring existing demand, while the founding team handled the operations required to serve it.

By June 2024, Local Kitchens announced another $40 million in funding, bringing its reported total to $75 million. The company then had 12 Northern California locations. It reported unit-level profitability and a model producing 50 percent more sales per labor hour than a traditional restaurant. That comparison was the company’s claim, rather than a universal measure of restaurant performance.

Goldsmith connected the efficiency argument to employees: more productive kitchens could support better wages and investment in training. That connection matters in a company asking one team to execute several different menus. The software cannot taste the result on a customer’s behalf. Training and execution remain part of the product, however elegantly an order moves across a display.

Leaving room for the next operator

The company’s language changed as it grew. The early description, “micro food hall,” gave way to “multi-brand restaurant.” Goldsmith’s earlier emphasis on bringing existing restaurant brands to new neighborhoods also developed into collaborations with chefs. The business was still concerned with variety, but its own identity increasingly had to carry the experience.

In May 2025, Local Kitchens announced Jay Gentile as CEO, with Goldsmith remaining chairman of the board. Goldsmith and Munday stepped away from daily operations. Gentile had already served as president and brought experience running restaurant businesses. Goldsmith publicly credited his ability to combine innovation with operational knowledge.

The subsequent Studio City restaurant showed where that work was heading: chef collaborations and a more developed experience for people eating on site. Goldsmith’s founding idea had become a restaurant business with decisions about service, menus, and dining rooms. A founder’s invention keeps meeting customers long after the founder hands over the daily schedule.

The people inside the system

Goldsmith describes himself as a learner and builder, interested in labor and operational excellence. His stated intellectual interests extend to generative anthropology, political philosophy, machine consciousness, and religion. It is an unusually broad reading list to place beside a business devoted to making the evening meal less troublesome.

Read together, those interests and his career suggest a useful way to understand his work: pay attention to the people inside the system. Tetra addressed the person trying to listen and take notes. Local Kitchens addressed the owner trying to expand, the cook preparing several menus, and the family trying to order together. Each project makes sense at the scale of an ordinary inconvenience.

The restaurant venture also supplied a particularly concrete test. A customer can forgive an abstract idea for being unfinished. Dinner comes with a shorter deadline. Goldsmith’s place in this story is as the founder who accepted that deadline, assembled the team to work against it, and eventually handed the operating role to someone else. The people at the table can continue disagreeing about what to eat. At least the disagreement fits on one order.