Try Expect uses Lepton’s GPUs to train NLP models for real-time social media sentiment analysis
Scalable Compute for
Real-Time Analysis
Try Expect leverages Lepton’s NVIDIA Blackwell GPUs to analyze social media sentiment in real time, updating market odds and dashboards instantly—even during peak events.
Automated Trading
and Oracles
Try Expect uses Lepton’s Blackwell GPUs for real-time sentiment analysis, instantly updating market odds during high-traffic events.
Founding Team
AI product builders with
a proven track record
Dallin Bentley
CEO
Try Expect’s mission draws from Dallin’s experience at Buster, where he co-built Buster as an open-source platform for deploying AI data analysts – empowering everyone to explore data on their own. There, he saw firsthand the challenges of aligning technology with real user needs. Recognizing similar dynamics in prediction markets—where timely sentiment and data drive outcomes—he set out to build tools that enhance real-time decision-making through decentralized participation.
With Try Expect, Dallin aims to bridge the gap between crowd wisdom and actionable insights, empowering users to anticipate market shifts, election results, and more—transforming speculation into structured, collaborative forecasting.
Jett Hollister
CTO
Jetti is a data scientist and decentralized systems expert with deep experience in belief dynamics modeling, real-time predictive analytics, and distributed networks. His work on bsky-net—a temporal graph dataset tracking social connections, communications, and evolving beliefs—has honed his ability to benchmark and refine predictive models using real-world, streaming data. With expertise in processing social sentiment, news trends, and network dynamics, he excels at building systems that turn complex data into actionable forecasts. His background in decentralized platforms like Bluesky also equips him with the architectural knowledge needed to design resilient, scalable prediction markets. As a leader at the intersection of AI and Web3, Jetti is uniquely positioned to drive the development of a Try Expect that leverages real-time data, adaptive modeling, and decentralized infrastructure.