The most revealing trip in Erbil Karaman's product career began in a second-hand phone shop. He was working at Facebook, where the company's grand ambition to connect the world met a less grand reality: much of the world did not own the phones used in Silicon Valley demos. Karaman took his team to Africa, bought the sort of low-end devices people could afford, loaded them with data, and tried the product. Facebook struggled. Storage was tight. Data was dear. The experience that looked effortless in California became a stubborn blue weight in the hand.
It was an engineer's field lesson with a novelist's moral: the system always looks different from the cheap seats. Karaman's response was not to ask users to catch up. The product had to travel toward them. Meta's engineering team later credited him among the people who helped improve Facebook for Android as the app's size fell by 65 percent during that effort. The work also sat beside his role in Internet.org, Facebook's attempt to bring more people online.
Years later, Karaman is still studying what happens when an elegant system reaches an inconvenient life. The object has changed from a social app to money. The constraint is no longer a thin data plan. It is time - the expensive interval after a payment has been authorized but before the recipient can put the funds to work.
“The whole idea of decentralization isn't just a technical term.”Erbil Karaman
A career made of bottlenecks
Karaman studied computer engineering at Middle East Technical University in Ankara from 2003 to 2009. By the time he graduated, he was already working on SpeedDate, where he served as a founding engineer and growth lead. The company was later acquired by Match.com. He went on to co-found Sprongo, then joined Facebook in 2012. His roles there moved through growth, engagement, mobile adoption, Internet.org, Facebook's London office and new product initiatives.
The titles tell one story. The sequence of constraints tells a better one. SpeedDate was about reducing the distance between strangers. Facebook's emerging-market work wrestled with storage, bandwidth and device cost. Lyft, where Karaman later became a senior product leader and Head of Product for Consumer and Platforms, coordinated distance, supply and routing. At EarnIn, where he served as chief product officer, the awkward interval sat between labor performed and wages received.
Along the way, he kept another foot in the founder world. He mentored young entrepreneurs through the Thiel Foundation's fellowship, became an advisor to 500 Startups' Turkey-focused fund, and made early investments. When 500 Startups announced its Turkish team in 2015, Karaman spoke of an ambition to help Istanbul become a major technology hub. Former collaborators describe a builder who looks for root causes, dislikes wasted motion and keeps his energy during the untidy moments when a startup must change course.
The payment is done. The waiting is not.
Huma Finance grew from the world Karaman encountered at EarnIn, where he met Richard Liu, the former Google engineering leader who became his co-founder. EarnIn had made timing a consumer product problem: people can perform work long before the conventional payday grants access to their earnings. Huma first carried that logic into income-backed decentralized lending. Rather than ask borrowers to lock up more crypto than they wished to borrow, its early model let partners underwrite credit against income and receivables.
Then the canvas widened. In 2024 Huma merged with Arf, a company providing liquidity for cross-border payments. The combination pushed Huma more firmly toward what the team calls PayFi - payment financing built around stablecoins, real-world payment activity and onchain pools of capital. In September of that year, Huma announced $38 million in financing. Distributed Global led the equity portion, with participation from HashKey Capital, Folius Ventures, the Stellar Development Foundation and TIBAS Ventures. Part of the total represented investment in real-world assets on the platform.
The consumer sees a card approval as an ending. For a merchant or payments company, it can be the start of several days of settlement. Cross-border firms may keep capital parked in pre-funded accounts so recipients can be paid before the old rails finish their journey. That money is useful, but idle. Huma's pitch is to make short-duration liquidity available against real payment flows, allowing capital to arrive while the legacy machinery is still clearing its throat.
Where PayFi enters the picture
A simplified view. Underwriting, eligibility, risk controls and network mechanics vary by product and partner.
The underlying technology matters, but Karaman tends to explain it through ordinary frustration. A merchant has made the sale yet cannot use the proceeds. A remittance provider has promised speed to customers while its own money waits elsewhere. A payment business could grow faster if each dollar returned sooner. In an interview at Korea Blockchain Week in 2025, he described a goal of embedding Huma's infrastructure inside card networks, trade finance, supply chains and conventional payment systems so users would not need to understand the blockchain beneath it.
The user behind the abstraction
There is a philosophical line through Karaman's public remarks. He has said hierarchical systems never suited him easily. His attraction to web3 is partly an attraction to distributed ownership: a chance, in his formulation, for users to participate in the systems they animate rather than merely become inventory inside them. He is blunt about the pressure that public-market incentives can put on good intentions at large technology companies.
That skepticism is useful, provided it survives contact with the new system too. Onchain finance does not repeal risk, and a stablecoin does not make every borrower sound. Huma's work lives in underwriting, partner quality, duration, liquidity and operational details. The romance of instant settlement eventually meets the spreadsheet. Karaman's career suggests he knows this instinctively: ambition only becomes a product after somebody tests it on the awkward device, in the thin market, under the unfriendly conditions.
“Alternative solutions will close the gap and create new competition.”Erbil Karaman
His public life offers few flourishes of the celebrity founder. The memorable details are workmanlike: six internet ventures before Facebook, product teams built across countries, patents involving mobile data and transportation systems, advice given to younger founders. Colleagues have praised his ability to make problem-solving fun even in chaos. It is a more useful trait than visionary fog. Companies rarely fail from a shortage of dramatic adjectives. They often fail because the team has stopped enjoying the act of finding out what is broken.
Huma also returns Karaman to the geography that shaped his ambitions. Its network speaks to global trade, remittances and the uneven distribution of financial access. Its name comes from the Huma bird of Persian and Middle Eastern tradition, a creature said never to land. The metaphor risks being almost too neat for a company trying to keep money moving. Still, it fits Karaman's long-running concern: value should travel farther than the institutions that once controlled its route.
A founder's useful question
What can another builder steal from this story? Not a token design or a fashionable suffix. The transferable habit is to locate the delay users have mistaken for a law of nature. Buy the cheap phone. Follow the card swipe after the receipt prints. Ask where the wage sits between Tuesday's labor and Friday's account balance. Measure who pays for every pause.
It is also a lesson in choosing the right unit of observation. Companies measure transactions, users and revenue. Karaman's work repeatedly points toward a smaller unit: the frustrating minute between intention and result. Those minutes accumulate. A slow app consumes a data allowance. A late settlement consumes working capital. A distant payday consumes options. Once the hidden cost is named, product strategy becomes less theatrical and more exact. The team can ask whose time it is borrowing, what that delay costs, and whether software can return it.
Karaman's path from Ankara to Silicon Valley and then into decentralized finance looks varied on a résumé. Seen from the user's side, it is remarkably consistent. He keeps arriving at systems that function beautifully for the center and poorly at the edge. His answer is to move the product closer to the edge, where bandwidth thins, cash flow tightens and patience becomes a form of capital.
The bet at Huma is that money can learn something from the internet: open networks, faster movement, fewer geographic permissions. The harder task is making those ideals survive credit cycles, regulation and scale. Karaman has already seen what happens when a mission meets the machinery of a giant company. This time he is helping design the machinery.