Built for Production

• zkML: Production-ready zero-knowledge machine learning framework that enables fast, practical AI verification.
• Verifiable compute should run everywhere: Our proof system runs fast in browsers, IoT devices, and other constrained environments to meet the needs of rapidly evolving agents.
• Developer-first: Train models in PyTorch, TensorFlow, or any ML framework — export to ONNX and generate proofs with JOLT-Atlas. No circuit expertise required.
• Composed by design: Use verifiable compute and memory together so every retrieval and compute step carries a proof — from data ingestion to model output to payment authorization.

Use cases

• Spending policy proofs: AI agents prove they followed budget, vendor, and risk policies before USDC is released. No proof, no payment.
• Verifiable AI inference: Prove a specific model ran on specific inputs with a specific output — critical for compliance in regulated industries.
• Multi-agent workflows: Agent-to-agent handoffs with cryptographic receipts. Every query/response in the chain carries a proof.
• Privacy-preserving compliance: Prove policy adherence without revealing proprietary models or sensitive data. Selective disclosure for audit and regulatory requirements.

Pioneering Tech

• Scalable: Traditional bespoke circuits don't scale across diverse apps. That's why we built JOLT-Atlas — a zkML framework with 3-7x performance improvements over alternatives.
• Fast: Sub-second proofs for the classification models, embeddings, and guardrails that gate AI agent spending. Native ONNX support and optimized non-linear activations.
• No need for specialized hardware: zkML proofs are based on mathematics, not hardware trust. Unlike TEEs that depend on hardware, proofs run on any device — no vendor lock-in, no physical access vulnerabilities, deployable in-house for institutions that can't outsource trust to hardware manufacturers.

Our users

• Enterprises adopting provable AI memory and compute for security, auditability, and compliance.
• Agentic payment infrastructure — wallet providers, payment processors, and stablecoin platforms building verification into AI commerce.
• AI agents that need to prove they followed spending policies before transacting — autonomous systems generating their own cryptographic receipts to unlock payments.

• High-growth AI startups adding zkML to real products without overhauling their stack.

Founders: Wyatt Benno, Houman Shadab.