Case study

How Y Combinator Fights Fraud with SEON and AI

Company

Industry

Venture Capital

Talent Marketplace

Summary

Y Combinator (YC) operates Work at a Startup, the hiring marketplace that connects YC portfolio companies with talent in engineering, design, operations and more. As the volume and sophistication of fraudulent applicants grew, YC needed a way to screen applicants in real time, without slowing down legitimate ones. By integrating SEON across the hiring platform, YC built an automated, AI-driven line of defense that gets sharper with every applicant — and lets a small trust & safety team operate at the leverage of a much larger one.

Key Results

  • YC portfolio companies see fewer fraudulent applicants contact them
  • Substantially reduced trust & safety workload on bad-actor handling
  • A self-improving, AI-based system: every decision becomes training signal for the next applicant

About Y Combinator

Founded in 2005, Y Combinator has funded more than 5,000 startups, including Airbnb, Stripe, Dropbox, DoorDash, Reddit, and Coinbase. Through Work at a Startup, YC helps its portfolio companies hire the engineers, designers, and operators they need to grow.

“SEON came recommended to us by some of our top startups, and it’s delivered in every way we need. The product gives us a signal layer rich enough to build on — and AI lets us turn every applicant into a sharper model for further training and detection.”

Ryan Choi
Director of Work at a Startup, Y Combinator

The Challenge: A Hiring Marketplace at the Center of a Growing Fraud Industry

YC’s portfolio is filled with successful, fast-growing companies actively hiring. Over time, fraudsters targeted the platform with various hiring fraud schemes – ranging from fake candidate profiles to attempts to commit employment fraud. The Work at a Startup team needed to solve several problems at once: 

  • Identify synthetic profiles with AI-generated names, fabricated emails
  • Capture and integrate device signals and usage patterns that might indicate app abuse
  • Support a small trust and safety team to perform heavy manual review

“Work at a Startup should connect founders with real people. SEON helps us verify that accounts belong to actual humans before anyone reaches a founder’s inbox, so founders can spend their time evaluating candidates rather than filtering out bots and synthetic accounts.”

Erica Clark
Product Engineer, Y Combinator
Your AI Command Center for Fraud and AML

Fragmented tools create blind spots. SEON brings fraud detection, transaction monitoring and compliance workflows into a single view.

Speak with an expert

The Solution: A Unified Intelligence Layer Across Work at a Startup 

YC evaluated multiple identity and fraud-intelligence vendors before choosing SEON. The differentiator was the ability to consolidate signals that would otherwise come from a fragmented stack of point tools — alongside a configurable rule engine YC could tune to its specific applicant population, and a partnership model that scaled with the threat. 

Why SEON

  • Unified intelligence layer. A single integration replaces what would otherwise be a multi-vendor stack covering identity, device, and behavioral signals. 
  • A signal layer rich enough to build models on. SEON’s intelligence is the foundation; YC’s own models, learning from every decision, are the layer above. SEON specialists work alongside the team to evolve detection as the threat evolves. 
  • Configurable to YC’s population. YC tuned its detection logic to Work at a Startup’s applicant base — including protections for early-career applicants.
  • Bidirectional feedback loop. Every confirmed-fraud decision and every reversal is labeled back to SEON via the Transaction Label API, training detection on YC’s specific attack patterns. 

SEON is embedded across the full applicant lifecycle — from the first time a candidate lands on the platform, through profile creation, application, and ongoing activity. These signals are a critical input layer for YC’s own AI models — the data behind real-time identification and enforcement against fraudulent activity across the applicant journey on Work at a Startup. The richer the signal, the sharper the model. 

Results

Since rolling SEON out on Work at a Startup, YC has measurably hardened its hiring marketplace against industry-wide attacks on hiring platforms and companies: 

  • Fewer fraudulent applicants reach YC portfolio companies. Real-time screening catches the bulk of bad-actor traffic before it touches a founder’s inbox. 
  • Substantial reduction in trust & safety team workload. Routine bad-actor traffic is handled automatically, freeing the team to focus on policy work, ambiguous cases, and emerging attack patterns. 
  • A self-improving system. SEON’s signals and YC’s AI models sharpen each other with every decision.

Conclusion

The qualities that make YC startups attractive places to work make them attractive targets, too. SEON gives YC’s trust & safety team the foundation; YC’s internal AI systems use these signals to help thwart account abuse. The self-learning system gets sharper with every applicant — and every true positive. For YC’s founders, that means one thing: the applicants who reach them are the real ones.

SMG Cuts Investigation Time by 50%


SEON’s leading fraud prevention and AML compliance platform is powered by AWS.

Take the First Step Toward Transformative Fraud Prevention