Variance
Variance builds AI agents for fraud review, identity verification, and investigations, serving Fortune 500 companies, marketplaces, and regulated financial services firms with automation for the trust-and-safety workflows that traditionally required large manual review teams.
Visit Website ↗ + Add to CompareOverview
Variance’s AI agents automate fraud review, identity verification, and investigative workflows for large marketplaces and regulated financial services firms, targeting the trust-and-safety operations that traditionally require sizable manual review teams to keep pace with transaction and account volume.
Founded by Karine Mellata and Michael Lin — both previously responsible for Trust & Safety infrastructure at Apple and Discord — Variance went through Y Combinator’s Winter 2023 batch, is based in San Francisco, and has raised $3.1 million in seed funding (January 2024) followed by a $21.5 million Series A (March 2026).
Innovation Matrix Assessment
Progressing from a $3.1M seed round to a $21.5M Series A within roughly two years, while growing to a 12-person team, indicates solid and sustained execution pace.
Automating fraud review and identity verification workflows directly reduces the scale of manual review teams needed to handle growing transaction and account volume at marketplaces and financial services firms.
Raising a $21.5M Series A roughly two years after a $3.1M seed round is a strong funding-progression signal, indicating investors saw meaningful traction between rounds.
AI-driven fraud and identity review automation follows an established trust-and-safety automation trend rather than introducing a fundamentally new security paradigm.
No independent, third-party benchmark of fraud-detection accuracy or false-positive rate was found; efficacy evidence is not independently verified beyond the company's own claims.
Fraud and identity verification remain persistent, high-cost problems for regulated financial services and large marketplaces, keeping automation in this space durably relevant.
Why CISOs Should Care
Fraud and identity-verification review queues are a major operational cost center and a common source of both false-positive customer friction and missed genuine fraud; Variance's founders bring direct large-scale Trust & Safety infrastructure experience from Apple and Discord to that specific problem.
What Makes It Different
Variance's founder background running Trust & Safety infrastructure at hyperscale consumer platforms gives it operational credibility in fraud-review automation that most generic AI-agent vendors entering the space lack.
The Matrix Verdict
53/100 — INCREMENTAL INNOVATOR
A credible, well-funded fraud and identity AI agent platform with founders carrying directly relevant hyperscale Trust & Safety experience; scores reflect solid momentum and operational relevance for a Series A-stage company.
Editorial Note: Claims vs. Verified Findings
YC batch, founders, and both funding rounds ($3.1M seed, $21.5M Series A) with dates are independently confirmed via Variance's Y Combinator company page; specific customer names and efficacy metrics are not disclosed.
Sources
Alternatives to Variance
Adaptive Security
AI-driven platform that simulates deepfake, voice, and multichannel social-engineering attacks to train and test organizations against next-generation phishing.
Quilr
Early-stage agentic AI security startup building a 'Service-as-Software' platform to guard against human-related breaches and secure AI agent…
Zenity
Governance and security platform for AI agents and low-code/no-code development, securing agent identity, permissions and behavior across the…
Tenzai
An agentic AI penetration testing startup building autonomous 'AI hackers' to find and validate exploitable vulnerabilities at a…
Alice (formerly ActiveFence)
Israeli AI security company (rebranded from ActiveFence in January 2026) offering a lifecycle platform to test, guard, and…
Reco
Reco secures the "agentic ecosystem" — mapping what AI agents can access across SaaS and enterprise apps, detecting…