# Talent Protocol: 5 Years Building for Builders *What we built, what we learned and how talent.app became a public good.* By [talent](https://paragraph.com/@talent) · 2026-08-19 --- [IPTS](https://ipts.ai/) has acquired Talent Protocol, with support from [Protocol Labs](https://pl.xyz/). [View post on X (2090184097381470412)](https://x.com/protocollabs/status/2090184097381470412) Interplanetary Talent Services (IPTS) leads recruitment across the Protocol Labs network. The builder reputation we calculated will help its team find talent based on what they have built and contributed, for roles across AI, crypto, hardware, life sciences and other frontier technologies. You can read the [Protocol Labs announcement](https://pl.xyz/blog/ipts-acquires-talent-protocol/) for more context on the acquisition. This post tells the Talent Protocol side of the story: what we tried, what worked, what we learned and what happens now. After five years, Talent Protocol is ending commercial operations. The [talent.app](http://talent.app) is being replaced today by a lighter public-good version at [talentprotocol.com](http://talentprotocol.com). The team is not joining IPTS or Protocol Labs. We started with career tokens ----------------------------- Talent Protocol started in Lisbon on May 1st, 2021 with a question: what if people could invest in talent they believed in? Our first product let people create and buy career tokens linked to individual talent. The idea attracted thousands of people and generated roughly $500,000 invested in talent during its first months. Career tokens created real activity, but they also taught us that markets are imperfect judges of potential. The people who already had an audience usually attracted the most support. The market was better at measuring attention than finding talented people flying under the radar. We kept the original question and changed how we approached it. Instead of asking who people believed in, we started looking for evidence of what someone had built. Talent House made the mission real ---------------------------------- Talent House, later renamed Take Off, became the clearest expression of what we wanted Talent Protocol to accomplish. The first edition brought builders including Simon Puebla and Camelia to Barcelona. Over 12 editions, the program supported 116 builders from more than 15 countries with funding, mentoring, community and access to high-impact events. Protocol Labs was its main long-term sponsor, with more support from partners like Celo and ETHGlobal. One of those builders was [Mark Rachapoom](https://www.markrachapoom.com/), who received a full grant to attend Talent House Paris in 2022. Mark arrived from Thailand for his first crypto conference and described the group as an “amazing family.” He later founded Merse, which joined Y Combinator’s S24 batch, and now leads Dench. Mark did the hard work. Talent House helped put him in a room full of people and possibilities that would otherwise have been harder to reach. Talent House was expensive, difficult to repeat and never a business. It was also some of the work we are proudest of. Dashboards can count millions of profiles and transactions. Spending a week with young builders shows what access, encouragement and the right peers can change. ![](https://storage.googleapis.com/papyrus_images/14be70a526fc72098824e0647ff66b347306d0f83913e39ee7ba9c2944388b2a.jpg) Helping builders find the recognition they deserve -------------------------------------------------- In 2024, we launched Talent Passport and Builder Score. Builder Score brought together GitHub contributions, onchain activity, hackathon participation, credentials and other signals into a verified record of a builder’s work. It gave people another way to be seen beyond job titles, education or how well they promoted themselves. Builder Score was displayed on Basenames, Etherscan and used by many other partners as a key primitive for discovery, hackathons, grants and rewards. Over time, Talent Protocol indexed 6.6 million builder profiles and 86 million events. We built one of the largest structured records of what builders were working on and contributing to. People found the data useful. Turning that usefulness into a business was much harder. Most API users stayed on the free tier. Builders resisted paying for a score they might disagree with. Ecosystems liked having better information but rarely bought reputation as a standalone product. The data became more valuable when another product used it to do a specific job. [View post on X (1836070896433553913)](https://x.com/jessepollak/status/1836070896433553913) The closest we came to a sustainable business --------------------------------------------- Builder Rewards used the same reputation data to distribute funding based on verified work. Builders did not need to write applications or make claims. Their contributions were the application. ![](https://storage.googleapis.com/papyrus_images/0182b8fd2c23c80956dc35d6f6fb79686c8509eccb325a56a0dccf84d660d1c7.jpg) Base sponsored the first program. We later worked with ecosystems including Celo, WalletConnect, Syndicate and Stacks. In 2025, more than $600,000 was distributed to builders. For many recipients, the amount mattered less than the timing. Receiving early recognition for open-source work or a project that had not yet become a company could provide capital, confidence and an audience when all three were scarce. Builder Rewards gave us our strongest signs of product-market fit and became our best source of revenue. But it also exposed the limits of the market. The number of crypto ecosystems with large, recurring budgets for builder rewards was small. Those budgets usually came from marketing and community spending, which rose and fell with market cycles. We built a promising product, but its durable customer value was never fully proven before budgets began shrinking. In 2026 the crypto industry changed. More attention and capital moved back toward financial products. Builder reputation, community programs and public goods had a less obvious place in that direction, especially when they were not connected to a financial use case. Talent Protocol gradually lost its role in the industry it had been built to serve. Market conditions made the problem harder, but they do not fully explain the outcome. A sustainable company needs repeatable demand and revenue even when its market becomes less generous. We never reached that point. The final experiments --------------------- We kept looking for other ways to turn the data into a business. Talent+ was a paid monthly subscription for builders who wanted more from the Talent app. Talent Pro was a later B2B experiment that combined our builder data with AI models for recruiting. Talent Pro made sense on paper. Companies needed better ways to find technical talent, and we had a unique record of what millions of builders had done. In practice, we did not find enough demand to pursue it. We also learned that building a serious recruiting product would require starting almost from scratch, with new workflows, services and customer relationships. That experience helped us understand why IPTS is a better home for our deep builder record. IPTS already works directly with frontier-technology companies and understands how technical recruiting happens. The data can strengthen a product and service that already exists. Why the company had to close ---------------------------- By 2026, we had built valuable data, products and relationships. We had not built a sustainable business. There was not much runway left. The choice was no longer whether to stop, but how. Continuing until the funds ran out would have left less time and fewer options to close the business properly. We focused the remaining time on finding a new home for what we had built. Protocol Labs had been part of Talent Protocol’s story since the beginning. PL Capital was our largest investor. Organizations across the network supported our products, sponsored Talent House and worked with us over five years. The Protocol Labs network helped bring Talent Protocol and IPTS together for this acquisition. The acquisition preserves useful work and gives our builder data a new purpose. It also closes a company that never found sustainable revenue. We are proud of the first and honest about the second. What we learned --------------- ### 1\. Reputation creates more value when it helps perform a specific job A reputation score can be interesting on its own. It becomes useful when someone uses it to allocate funding, find a candidate, assess an application or discover work they would otherwise miss. The same logic now applies to IPTS: the reputation data can help find builders for real opportunities. ### 2\. Product demand and market size are separate questions Builders loved receiving rewards. Ecosystems paid us to run programs. That gave us promising signs that the product solved a problem. But the market still too small and too dependent on a limited number of crypto budgets. A product can work and still lack enough customers to support the company behind it. ### 3\. Markets measure attention as well as potential Career tokens taught us this early. Open markets made support visible and gave people a way to back someone they believed in. They also favored people who already knew how to attract attention. Builder Score reduced some of that bias by using evidence of work. It still measured what someone had already done better than what they might do next. Predicting human potential remains a much harder problem. ### 4\. Revenue deserves earlier tests We spent years proving that people cared about builder identity, reputation and recognition. We should have tested willingness to pay with the same intensity and much earlier. Free adoption can teach you a lot about a product. It cannot tell you whether the company will survive. ### 5\. Individual stories matter The 6.6 million profiles show the size of the record. The $600,000 shows how much funding moved through Builder Rewards. The 2,000+ builder names displayed in Times Square show how large the community became. Mark’s story explains why we cared. The moments we remember are builders getting on their first flight to a conference, meeting onchain friends, receiving their first reward or seeing their work recognized publicly. Those outcomes were smaller than the company we hoped to build. But they were real and impactful. ![](https://storage.googleapis.com/papyrus_images/93b48109d2662bc3be5f52d526573107925f7b907e87b593b5314637fd3858f6.jpg) Talent House in Rio de Janeiro Talent Protocol is now a public good ------------------------------------ Talent Protocol’s commercial products and company are closing, but [talentprotocol.com](http://talentprotocol.com) is staying online as a public good, with a new version of Builder Score built to work without the company behind it. ![](https://storage.googleapis.com/papyrus_images/779771bb82191832158a402798316ea596095a87cc28f8d629ff85f71d6d82e0.png) [talentprotocol.com](http://talentprotocol.com) as a public good We moved the core of Builder Score into the browser. Enter a wallet address or ENS name, and optionally a GitHub handle, and the app gathers public data from 4 chains and GitHub. It computes the score locally using open rules and shows the formula behind every point. The result is a score you do not have to trust. Anyone can inspect the inputs, rerun the calculation and verify it. Builders can also publish their score as an EAS attestation on Base, creating a public onchain record that others can verify by recomputing the score. The app uses Base Builder Codes, allowing Base to attribute each attestation to Talent Protocol. The public-good version has no accounts or central database to maintain. It is designed to keep working without Talent Protocol operating it. The `$TALENT` token was not part of the acquisition, and there will be no new product or token roadmap. Only one scenario remains: if Base launches a token and considers activity attributed through Builder Codes when deciding allocations, Talent Protocol could receive an allocation. If that happens, `$TALENT` ownership could be used to determine eligibility for a future `$BASE` airdrop. What happens to the rest of the company? ---------------------------------------- IPTS will receive a selected subset of profiles with promising builder records and available contact information. This will help those builders become visible for relevant opportunities across frontier technology. The final Builder Rewards campaign, Celo Proof of Ship, ended on 27 July 2026, and rewards were distributed on 3 August. The API, Talent+ and Talent Protocol’s social channels will close at the end of August. Until 31 December 2026, questions and data requests can be sent to [contact@talentprotocol.com](mailto:contact@talentprotocol.com). Ecosystem Rewards will end on 29 October 2026 with a final distribution and burn. Existing Sablier streams are decentralized and will continue until 29 November 2027. The deal with IPTS did not include Talent Protocol’s legal entities, customer contracts or team. It did not create sale proceeds to distribute. The remaining company funds will cover outstanding obligations and help us close the business properly. Thank you, builders! -------------------- Talent Protocol existed because thousands of people believed there could be a better way to recognize and support builders. Thank you to every builder who created a profile, connected an account, launched a career token, applied to Talent House, used Builder Score or received Builder Rewards. Thank you to our investors for backing a difficult problem before the path was clear. Thank you to Protocol Labs for being there from the first funding round through the final acquisition. Thank you to Base, Celo, ETHGlobal, WalletConnect, Stacks, Coinbase Ventures, Subvisual, and the many other organizations that funded builders, tested products and trusted our team. And thank you to everyone who was part of the Talent Protocol team: Andreas, Fred, Guedes, Gustavo, Isabella, Ivan, Juampi, Luminita, Panos, PP, Ruben, Rudnevsky, Sara, Simão and Tolga. A small team built career markets, scholarship programs, identity products, scoring systems, reputation APIs, reward infrastructure and a public record covering millions of builders. The company changed direction many times. The team kept building. Talent Protocol was led throughout these 5+ years by its co-founders Pedro Oliveira, Filipe Macedo and Francisco Leal. We are grateful to everyone who trusted us with their work, time and belief. ![](https://storage.googleapis.com/papyrus_images/338f15cdfa20bb153c4bffcdd00e3ab91660aadb56fd2adc428527a6e35ce380.jpg) \-- **Pedro Oliveira, Filipe Macedo and Francisco Leal** Co-founders, Talent Protocol --- *Originally published on [talent](https://paragraph.com/@talent/talent-protocol-5-years-building-for-builders)*