Breaking patternCheckout was the easy partReturns, claims and invoices enter the chatAudrey Djiya builds for the work behind the orderBreaking patternCheckout was the easy partReturns, claims and invoices enter the chatAudrey Djiya builds for the work behind the order

Founder profile / Operations, with receipts

Audrey Djiya Is Fixing Everything That Happens After You Click Buy

The checkout page gets the applause. Audrey Djiya built Handled for the lost shipments, return limbo and spreadsheet archaeology that begin after it.

Buying online is now an act of magnificent laziness. A thumb moves. A package appears. Between those two events sits a small republic of software, warehouses, carriers, support agents and billing systems, all expected to agree about what happened. Frequently, they do not. The parcel pauses in New Jersey. The return loses its paperwork. The inventory count develops a private theory of reality. Somewhere, an operator starts opening tabs.

Audrey Djiya has made a company out of that somewhere. Handled, the New York operations platform she co-founded and now leads, works on what the industry calls post-order operations: tracking shipments, resolving exceptions, processing returns, filing claims and adjusting invoices. These are necessary verbs with no glamour budget. They are also the point at which a promise made on a polished storefront collides with a loading dock, an address typo or a carrier scan that never arrived.

Handled connects the systems that already hold fragments of the story, from commerce and warehouse software to carriers and customer support. It is meant to give brands and third-party logistics providers one view of orders, shipments, returns and inventory, then execute the follow-up work when a problem appears. The proposition is refreshingly testable: fewer humans should have to become detectives every time a box wanders off script.

“As order volume picked up as these brands grew, operational chaos scaled alongside that growth. Teams were just drowning in manual work.”Audrey Djiya, on the problem behind Handled

Learning to hear the machinery

Djiya’s route to this problem began long before anyone put “AI operations” on a landing page. She was born in Queens to parents who had immigrated from Cameroon, then grew up mostly in Northeast Pennsylvania. Home remained distinctly African, she has said, and frequent trips to Cameroon made the continent’s business life feel personal rather than theoretical. She watched her parents try to start businesses. At Dartmouth, where she graduated in 2017, she studied government and international studies with a focus on sub-Saharan Africa.

There was another curriculum running alongside the official one. Djiya had taken voice lessons since age five, performed in musicals and choirs, and became president of the Dartmouth Decibelles, an a cappella group. Group singing is an education in systems disguised as entertainment: individual parts, shared tempo, constant listening, no useful place to hide. It is tempting to over-neaten a life in retrospect, but the image suits her later work. Operations also depends on separate actors staying in time.

After Dartmouth she joined Deloitte’s public-sector practice, working largely with nonprofit clients in Washington and gaining exposure to entrepreneur accelerators in Kenya and Tanzania. She contributed research and analysis to a Deloitte Insights study on service design in government. The phrase sounds bureaucratic. The underlying idea is intimate: begin with the person who must actually use the thing.

2017Dartmouth graduate, government and international studies
S24Y Combinator batch for the company first known as Zimi
2025Forbes 30 Under 30, Retail & Ecommerce

The voice of the customer gets a passport

Djiya moved from consulting into American startups, first at wedding company Zola and then Peloton. At Zola, her title was customer insights manager. Her description was better: “I serve as the voice of the customer across the organization.” Surveys went out after signup and after weddings, moments when people were especially able to say where the product delighted or annoyed them. It was structured listening, timed to experience rather than executive convenience.

Peloton gave her experience with strategy and high-growth commerce. It also sharpened a habit she would later have to revise. In a 2021 interview, Djiya described being accustomed to months of user research before a launch. That thoroughness is admirable until the market moves while the research deck is still choosing a font.

She wanted to work in African technology but found an old circular requirement: companies wanted local experience, while getting local experience depended on already having it. Venture for Africa offered a way through. The fellowship paired her with Carbon, a Nigerian fintech company evolving from a lender into a broader digital bank. Djiya was drawn to the reach of financial products and the practical stakes of inclusion. Money, as she put it, affects everyone.

At Carbon, data became less a report than an argument. Djiya helped lead the implementation of Heap Analytics, which showed where users dropped away and where app flows created friction. The team found that its signup design discouraged completion and iterated on it. She also worked on the in-app rollout of Carbon Zero, its buy-now-pay-later product. The lesson was not to worship speed. It was to make speed answerable to evidence.

“The first version doesn’t have to be the best version - it’s just an MVP, and as long as you’re looking at data and learning from your users, it can only go up from there.”Djiya, reflecting on her time at Carbon

That sentence contains most of a product philosophy: humility, measurement and permission to revise. Djiya stayed in African fintech, later serving as a senior product manager at Duplo. At Stanford Graduate School of Business, where she earned her MBA, she continued researching commerce in African markets, including fieldwork on social selling in Nigeria and Senegal. Each stop added a different lens: the customer interview, the operating model, the product funnel, the local context.

Audrey Djiya and Peter Nsaka standing beside a Y Combinator sign outdoors
Summer school for founders: Audrey Djiya and Peter Nsaka at Y Combinator, where their company joined the S24 batch. The sign is tidy. Startups generally are not.

A company changes its address

In 2024, Djiya co-founded Zimi with Peter Nsaka. The original company tackled cross-border commerce, helping international merchants place inventory closer to American buyers while managing fulfillment and payments. The concept joined Y Combinator’s Summer 2024 batch and earned Djiya a place on the Forbes 30 Under 30 list for Retail & Ecommerce in 2025.

The company’s public identity later became Handled, and the focus narrowed toward a problem Djiya had observed inside high-growth brands: operational chaos did not merely accompany order volume; it grew with it. Every additional sale could generate another exception, another support request, another manual reconciliation. More growth meant more people hired to keep mismatched systems in conversation.

By early 2026, Djiya was leading Handled as its sole remaining founder. The evolution from Zimi’s cross-border platform to Handled’s post-order automation looks less like abandoning a subject than moving deeper into it. Both businesses live behind the storefront. Both ask why selling across systems and borders still requires so much human translation. The new version simply names the labor more plainly.

The post-order relay
Unify order, warehouse, carrier and support data
Watch for delays, mismatches and return events
Execute claims, adjustments and follow-up work
Give operators one current view of what happened

The useful kind of artificial intelligence

Handled arrives during a season when almost every software company is tempted to introduce itself as artificial intelligence wearing a respectable jacket. Djiya’s framing is more concrete. In a recent note to brands attending Shoptalk, she listed purchase orders managed over email, sales orders manually entered into enterprise software, stockouts predicted by one person’s spreadsheet, returns in limbo and lost shipments discovered by customers. The list has the charm of a kitchen inventory after a small fire. It also makes the product legible.

Automation here is not a chatbot floating beside the work. It is supposed to perform the work: notice a shipment delay, process a return, file a claim, adjust an order or correct billing. That distinction matters. A system that drafts a polite paragraph about a missing package may be impressive. A system that resolves the missing package is useful.

Third-party logistics providers feel this pressure acutely. They sit between brands, warehouses, carriers and customers, inheriting the contradictions of each. On The New Warehouse podcast in February 2026, Djiya discussed the move from reactive firefighting to proactive execution. Firefighting is exciting in films and expensive in operations. The best incident, after all, is the one the customer never needs to report.

CONSULTING

Public-sector service design and work with entrepreneur programs in East Africa.

COMMERCE

Customer insights at Zola and strategy at Peloton.

FINTECH

Product analytics at Carbon, followed by product leadership at Duplo.

FOUNDING

Zimi enters YC S24, then evolves into Handled’s post-order operations platform.

The pattern worth stealing

Founders are often advised to find a pain point, a phrase so scrubbed of feeling that it can mean nearly anything. Djiya’s career offers a more useful version: follow the repeated annoyance across contexts. At Zola, listening exposed where people struggled. At Carbon, analytics showed the precise point of friction. At Peloton and other commerce businesses, growth revealed how manual operations compound. Handled turns those observations into a single wager.

There is also a lesson in how Djiya crossed domains. Government consulting, wedding registries, connected fitness, Nigerian banking and warehouse automation do not resemble a conventional apprenticeship. Yet each taught her to watch people interact with systems that were designed elsewhere. The founder’s advantage was not knowing one industry from birth. It was developing a method for entering messy rooms: ask, observe, measure, build, revise.

Djiya once said she loves “the process of creation and the ways that technology can be used to make people’s lives better or easier.” The phrasing is modest, almost suspiciously so in an industry that prefers to revolutionize breakfast. Easier is a serious ambition. Easier means the warehouse manager goes home without checking one more carrier portal. It means the support agent answers with certainty. It means a growing brand does not add a new spreadsheet each time sales rise.

The checkout page will continue to receive the better photographs. Fine. Djiya is building for the ungainly middle, where a purchase becomes a promise and a dozen systems must keep it. Commerce has spent years perfecting the click. Handled’s bet is that there is a large, durable business in making everything after it feel just as simple.