$200 GPU credits for yctierlist foundersClaim yours →
Oddpool

Oddpool

Institutional data layer for prediction markets

Spring 2026ActiveB2BEngineering, Product and DesignFintechFinanceAnalyticsSan Francisco, CA, USA
Company
https://www.oddpool.com
Oddpool is a platform that helps quant funds and trading desks harness the power of prediction markets. We make it easy to access, search, and analyze every venue from one place. Avi and Ritesh have been roommates since freshman year at Georgia Tech. Avi has a Master's in ML and saved Microsoft $6M/year by building a more accurate model on half the hardware. Ritesh has a Master's in Distributed Systems and made buy-side trading systems at Bloomberg 40% faster. We started trading prediction markets ourselves but couldn't backtest our strategies because the tooling for this emerging asset class didn't exist. The deeper we went, the more we saw that fragmentation between prediction markets and lack of clean data was going to be a big problem. We launched a prosumer tier and grew to several thousand dollars in MRR within months. Now we're bringing it to institutions, building tooling with the rigor and depth serious desks demand. They're the fastest growing asset class in finance, and CFTC applications for new exchanges keep climbing. Each new venue makes the market bigger, the fragmentation worse, and the unified data layer more valuable. It doesn't exist yet. We're building it.

Note: This is a preliminary assessment based on limited publicly available information. We did not have access to LinkedIn profiles or live product screenshots for this analysis. We will update this entry with a more thorough review soon.

Verdict

High Signal
Market Opportunity
Prediction markets are the fastest-growing asset class in finance with active CFTC exchange applications accelerating fragmentation. B2B ICP is quant funds and trading desks — high-value, high-willingness-to-pay buyers who need institutional-grade data infrastructure. Data and analytics tools for financial markets command premium SaaS pricing; comps like Bloomberg terminal subscriptions validate the monetization model.
High Signal
Founder Signal
Avi has an MS in ML from Georgia Tech and built a prompt injection model at Microsoft securing 20B requests/year while saving $6M/year — shipped real production ML. Ritesh has an MS in Distributed Systems and at Bloomberg architected trading infrastructure processing 100K+ TPS — directly relevant to the institutional data product they're building. Both have graduate-level technical depth and domain overlap with the exact problem they're solving.
Medium Signal
Competition
No direct competitor data surfaced, which is consistent with their claim that unified institutional tooling for prediction markets doesn't yet exist. Indirect competition includes general market data platforms (Bloomberg, Refinitiv) and vertical prediction market analytics startups, but none appear purpose-built for cross-venue prediction market backtesting at institutional depth. The risk is that this niche may stay small if prediction markets don't achieve mainstream institutional adoption.
Medium Signal
Product
Has a live prosumer tier with 'several thousand dollars in MRR within months' — real early traction, not vaporware. However, no named customer logos, no pricing page visible, no API docs cited, and no specific institutional clients confirmed yet. The core product (unified data layer + backtesting for prediction markets) addresses a real gap they personally encountered.
OverallA Tier

Two technically deep co-founders with directly relevant production experience at Microsoft and Bloomberg, building in a real gap they discovered through personal use — strong founder-market fit. Early MRR from a prosumer tier validates demand before the harder institutional sales motion kicks off. The main risks are: (1) prediction markets remain a niche asset class and institutional adoption stays slow, limiting TAM realization, and (2) the transition from prosumer SaaS to institutional sales requires a very different GTM motion neither founder has explicitly demonstrated. But the technical credibility, the genuine market fragmentation problem, and early revenue make this a credible A-tier bet.

Active Founders

Avi Arora
Avi Arora
Founder

Founder, CEO @ Oddpool | Prev. ML Researcher @ Microsoft Trained the prompt injection model at Microsoft which secures 20 billion requests / year CS @ Georgia Tech

Ritesh Malpani
Ritesh Malpani
Founder

Founder, CTO @ Oddpool | Prev. SWE @ Bloomberg Architected systems processing 100K+ TPS across trading infrastructure CS @ Georgia Tech

Oddpool
Oddpool
TierA Tier
BatchSpring 2026
Team Size2
StatusActive
LocationSan Francisco, CA, USA
Disagree with this rating? Let us know
YC
$200 GPU creditsClaim yours →