How Eragon hires world-class researchers and sources expert verifiers for post-training
Eragon runs two of its hardest talent and data problems on Clodo: finding world-class researchers to hire, and sourcing the niche expert verifiers it relies on for post-training and RL.
Eragon is an applied AI lab building an agentic AI operating system for the enterprise: instead of buttons, dashboards, and dozens of separate SaaS tools, teams run their business through a single natural-language prompt. Eragon connects a company’s existing tools and data, post-trains open-source models like Qwen and Kimi on its datasets, and deploys them inside the company’s own cloud, so each customer owns its own model weights and intelligence. Founded in 2025 by Josh Sirota and backed by a $12M seed round, Eragon is already in use across a handful of large businesses and dozens of startups.
The resultsEragon’s researcher hiring and verifier sourcing on Clodo, by the numbers.
- Researcher hires sourced through Clodo3
- Expert verifiers sourced8
- Hours saved per month19
The situation
Eragon is building an operating system for intelligence in the enterprise. Instead of buttons, dashboards, and a dozen disconnected SaaS tools, a company runs its business through a single prompt, on models it actually owns. Eragon connects a company’s tools and data and post-trains smaller, open-source models like Qwen and Kimi for the niche tasks and workflows that generic frontier models never see, deployed inside the customer’s own cloud, on the customer’s own weights.
That is a hard bet, and it only works with exceptional people behind it. Building purpose-built intelligence for every customer takes world-class researchers who can push post-training and RL, plus the niche expert verifiers who can judge whether a model’s output is actually right in a narrow domain. Both are the scarcest kind of technical talent, and neither falls out of a job-title filter.
The right researcher or verifier is defined by what they’ve actually done, the papers they’ve published, the systems they’ve built, the domains they can judge, not by a current title at a known company. That is exactly the kind of person generic sourcing tools and stale databases miss.
“By far the best people search I’ve ever seen. Like Exa but it actually works.”
Hiring world-class researchers
Eragon describes the researcher it wants in plain language, by research area, the problems they’ve worked on, the labs and papers behind them, the same way it would brief a hiring committee, and Clodo returns a ranked, enriched, contactable list of people who genuinely fit.
That turns weeks of digging through papers, GitHub, and profiles into a single search, so the team spends its time talking to strong candidates instead of hunting for them.
Sourcing expert verifiers for post-training & RL
Post-training and RL are only as good as the experts who verify the data. Eragon uses the same sourcing motion to find credentialed domain experts who can judge and label work in narrow, technical fields, the people qualified to say whether an answer is actually right.
Instead of settling for whoever is easy to reach, Eragon reaches the exact specialists each domain demands, which is what keeps the signal in its training data high.
One engine for talent and data
Hiring researchers and sourcing verifiers run on the same plain-language motion, so Eragon stands up either one without rebuilding its process or adding research headcount. Signal monitoring keeps both pipelines fresh, surfacing the right people as they publish, ship, and move.
Find and reach your target persona at the speed of thought.
