Lead, Market Deployments
Seer.pm
New prediction-market use cases, fundraising and business development.
Devansh
Mehta
Building mechanisms to solve the prisoner's dilemma, Goodhart's law, the tragedy of the commons and the free-rider problem to give a voice to the voiceless and speak truth to power.
01
About
Devansh Mehta started his career in 2017 as an investigative reporter, writing about U.S. military contracts awarded to an airline controlled by Azerbaijan's first family.
He then worked in Central India covering the Maoist civil war, where his company VoiceDeck - incubated at Microsoft Research India - deployed technologies for distributing news in villages through Bluetooth, collecting tribal-language data, running citizen journalism with the WhatsApp API, supporting crisis relief for migrants during the COVID-19 lockdown, paying people to learn public-interest information, and building an impact certificate marketplace.
He joined the Ethereum Foundation in 2025 to lead work on deep funding, a mechanism to scale impact evaluation by creating prediction markets for participatory budgeting. He currently leads market deployments at Seer.pm.
02
Impact initiatives
Most skilled at taking ideas from zero to one.
From a pure concept to initial empirical data, and finally to the learnings that arise from contact with the field. Devansh has gone through the loop of design → development → deployment → documentation ten times. See the complete research record on Google Scholar and a summary below.
| Project | Papers & field writing | What was learned |
|---|---|---|
| 01Deep Funding to improve and scale impact assessment |
| A funding mechanism that asks models and market participants to predict how expert juries will value open-source work. Human comparisons provide ground truth, prediction markets reward calibrated judgment, and the resulting weights can make grant allocation more scalable while exposing manipulation or noisy evaluation. |
| 02Validator Redirected Revenue |
| Contributed to the formal specification of a protocol-level mechanism that lets Ethereum validators redirect a collectively chosen share of staking rewards toward public goods. A majority-triggered contribution would apply to every validator, removing the ability to defect and addressing the free-rider problem. |
| 03Impact Evaluators Research Retreat |
| Organized a two-week research retreat in Reykjavík, Iceland, for 20 participants around Protocol Labs' Generalized Impact Evaluators framework. Its impact report records 22 open-access papers and $2.8 million in catalyzed grant funding, including work on Validator Redirected Revenue, Hypercerts v2, an academic grants round with AI-simulated judges, and Deep Funding's prediction market for distributing $350,000. |
| 04Arbitrum DAO Treasury and Sustainability Working Group Lead |
| Led the STEP program, diversifying roughly $50 million from the Arbitrum DAO treasury into stable, yield-bearing real-world assets while helping major providers launch on Arbitrum and building an endowment. The working group also researched CDPs and the ARB token, managed deliverables from karpatkey, Aera and Avantgarde, and helped develop a token-swap proposal for native projects that the DAO did not approve. |
| 05Impact certificates for better philanthropic fundraising |
| A pay-for-success marketplace that lets donors fund verified outcomes after they happen. The work translated impact certificates from a conceptual coordination mechanism into a live journalism prototype, testing whether tangible civic outcomes can be sold as a product to customers, similar to how one may buy a T-shirt. |
| 06Revitalizing the tribal Gondi language | Community-led collection turned scarce Gondi language data into dictionaries, children's stories, an IVR awareness service and a Hindi-Gondi machine translation model. The deployments produced more than 60,000 translations and engaged hundreds of community members in taking their language online. | |
| 07Mobile Satyagraha to connect rural and urban communities | A system for urban volunteers to adopt problems reported by rural citizen journalists and pursue resolution with public authorities. During India's lockdown it helped triage hundreds of distress calls, while the post-deployment research documented both real relief outcomes and the operational risks of entering crisis response unprepared. | |
| 08Using the WhatsApp API for rural outreach | A bridge between a long-running toll-free voice forum and WhatsApp, designed for communities gradually moving onto smartphones. Villagers could submit stories in audio, video or text, editors could verify them, and the newsroom could return published reporting through a familiar, scalable channel. | |
| 09Learn2Earn: paying people to learn information |
| An IVR system that rewarded people with small mobile-airtime payments for listening to public-interest messages, answering knowledge questions and sharing campaigns. Field deployments tested the idea across domains such as HIV education in Solapur, voting rights during the Indian national elections, and awareness of government programs. |
| 10Spreading information in media-dark zones using Bluetooth | An Android distribution model for places without reliable internet or cellular signal. Local participants carried audio news into villages and shared it phone-to-phone over Bluetooth, with incentives designed around the role of a digital newspaper carrier. |
03
Curriculum vitae
A career spanning investigative journalism, frontier technology and funding mechanisms.
Open full CVEducation
Columbia University
MIA, SIPA & MS, Journalism
St. Stephen's College, University of Delhi
BA (Hons), Philosophy
Lead, AI x Public Goods & Governance
Ethereum Foundation
Led Deep Funding and experiments in model-assisted public-goods allocation.
Founder-Director
VoiceDeck Foundation
Outcome-based financing for journalism and civic impact.
Founder-Director
VoiceDeck Private Limited
Five field technologies reaching more than 10,000 people in rural India.
Stabile Investigative Fellow
Columbia Journalism School
Investigated Pentagon contracts and U.S. loan guarantees linked to Azerbaijan's ruling family.
