A personal-finance dashboard can tell you that a subscription renewed three times without being used. It can sort the charge, compare it with last month and paint the waste in a handsome shade of red. Then it stops. The cancellation page, the password reset and the faintly embarrassing chat with customer support remain your problem. Ali Tabba has built his new company around that last, stubborn mile.
Gravy, the San Francisco startup Tabba co-founded with Max Church, is meant to move from observation to execution. The company describes an agent that can connect to banks, brokerages and email, then carry out instructions in the background as a user's finances change. The launch examples are deliberately domestic: track an angel investment mentioned in Gmail, cancel a recurring subscription, organize credit cards, prepare tax records, or move money to cover rent and bills.
The pitch sounds simple because the chore is familiar. The product problem is harder. Money software has spent years teaching people to look at graphs. Gravy asks them to grant software enough context, and eventually enough authority, to do something useful. Tabba's path to that question runs through grocery delivery, automated job applications and a large asset manager. Each stop looks different from a distance. Close up, they share a fascination with the handoff between seeing work and finishing it.
01 / Early velocityThe company before the famous company
Tabba was 18 when he co-founded KiranaKart in 2020. The venture sits in startup history as the company that became Zepto, the Indian quick-commerce business built around ten-minute grocery delivery. That shorthand can flatten the people who passed through the earliest version, but it does establish an important fact about Tabba: before he had finished university, he had already worked on a consumer operation where time was the product.
Quick commerce turns ordinary friction into an engineering and logistics problem. A grocery order is not glamorous. It is a list, an inventory system, a picker, a route and a promise. The customer's experience is the compression of all those moving pieces into a short wait. Tabba did not remain in the operating team that scaled Zepto, but he stayed a shareholder until 2024. His biography carries the early chapter without claiming the later company's execution as his own.
He studied economics at the University of Warwick, after school at Karachi Grammar School and Dubai College and an MIT LaunchX entrepreneurship program. The sequence matters less as a pedigree than as a clue to the two languages he would keep mixing: how people make choices and how products reorganize those choices.
02 / A useful reunionThe co-founder relationship that survived the first product
In late 2022, Tabba started Piko with Church. The product promised to reduce a repetitive job search to a single click, automatically filling applications and drafting cover letters. An introduction from the Initiator VC student founder program gave Tabba one of his few public quotes: “We are working hard to transform how people apply to jobs and internships going forward.”
Piko did not become the pair's permanent company. It produced something more portable: evidence that they could build together. Church, a technical founder who later led AI engineering work at Moonpig, returned as Gravy's chief technology officer. Tabba returned as chief executive. Repeat co-founders have an advantage no pitch deck can manufacture. They have watched each other deal with the dull Tuesday after the exciting launch.
“We are working hard to transform how people apply to jobs and internships going forward.”Ali Tabba, introducing Piko in 2023
The connection between Piko and Gravy is not a shared market. It is a shared unit of value: a task someone would prefer not to perform. Job applications contain forms, repeated fields and small adaptations. Personal finances contain statements, renewals, receipts and small decisions. In both cases, the software becomes interesting when it crosses from generating text into manipulating a workflow.
03 / The finance interludeLearning where consequences live
Tabba joined BlackRock as a summer analyst and then worked as an analyst in its iShares business from 2023 to 2024. It was a short corporate chapter between startups, but it placed him inside the machinery of investment products. Consumer software rewards ease. Financial software also has to account for risk, explanation and the possibility that a seemingly minor action has a real cost.
That tension is visible in Gravy's architecture. The product is presented as local-first, with sensitive financial data stored on the user's Mac. It can bring accounts into one view and accept natural-language requests, while keeping the underlying records on the device. Privacy here is not decorative copy beside a padlock icon. It shapes where the product runs.
Show a balanceInterpret
Explain a patternAct
Move, cancel, file
The local-first decision does not dissolve the trust problem. An agent that can act must still be predictable. It needs clear permissions, reversible moves and moments where a person remains in control. But keeping the data close to the user turns one abstract concern into a tangible boundary. It says the system can become more capable without making a remote server the default home of a person's financial life.
Do not begin with the dashboard. Begin with the postponed task. Then design the smallest permission that lets software finish it safely.
04 / Gravy's wagerFrom “what happened?” to “please handle it”
Gravy entered Y Combinator's Spring 2026 batch as a two-person San Francisco company. Its public line is unusually concrete: “Agents that move money, trade stocks, and file taxes.” The launch description widens the picture. The agent can watch accounts, investments and messages, build custom pages around the user's questions, and handle background work as new information arrives.
A traditional finance app has a fixed information architecture. Someone at the company decides that spending belongs in one tab, investments in another and credit cards in a third. Gravy proposes a more personal interface. Ask it to track several cards with a debt-payoff plan and a points check, and it can assemble a page for that job. Ask it to follow rent, deposits and utility bills from a rental contract, and the workspace can form around that instruction.
This is where the product becomes more than chat laid over a ledger. The interface can be generated around intent, while the agent remains present as the numbers update. The idea is less like visiting a banking website and more like leaving a clear note for a diligent financial assistant. Whether users will grant that level of access depends on the quality of the boundaries as much as the quality of the model.
“In 5 years, the way everyone manages their money will look radically different.”Gravy's 2026 launch statement
05 / The patternA career built around removing the handoff
It is tempting to narrate a young founder's career as a sequence of wins. Tabba's record is more useful when read as a sequence of attempts. Voltap came before KiranaKart. Piko came before Gravy. He crossed from startups into BlackRock and back again. Some chapters were brief. One became associated with a company far larger than its original form. Another ended with the same two founders choosing to try together once more.
The pattern is not scale. It is compression. KiranaKart worked on the wait between wanting groceries and receiving them. Piko worked on the repetition between finding a job and submitting an application. Gravy works on the gap between understanding a financial task and completing it. The tools change. The impatience with administrative drag remains.
That is also why Gravy's mundane examples do so much work. Cancelling a forgotten subscription is not a grand theory of finance. It is measurable. The user can see whether the agent succeeded. Small chores create the receipts from which trust can grow. Only then do larger actions, from moving money to preparing taxes, feel like a reasonable extension rather than a speculative promise.
Tabba is now building at the meeting point of his earlier worlds: the consumer speed of quick commerce, the workflow automation of Piko, and the financial consequence of BlackRock. Gravy's future rests on turning those ingredients into a habit people are comfortable leaving open on their laptops. The company does not need to make money management dramatic. It needs to make the neglected task disappear, and leave behind a clear record of what happened.