Asymptote Labs
Asymptote Labs builds runtime infrastructure and a shared data layer for AI agents. It gives security, engineering, platform, and FinOps teams real-time visibility, governance, detection, and control across agents, tools, users, and environments.
At a Glance
- Enterprise security and IT teams
- Engineering and developer-platform teams
- AI FinOps teams
- Regulated enterprises
- +1 more
AI Tools by Asymptote Labs
(1)
Beacon
Open Source Memory for AI Agents
Discussions
No discussions yet
Be the first to start a discussion about Asymptote Labs
Latest News
Asymptote Labs raises an oversubscribed $5.5M seed round led by Tower Research Ventures
Asymptote achieves SOC 2 Type II compliance
Beacon adds ChatGPT browser-chat telemetry
Beacon adds Claude.ai browser-chat telemetry
Products & Services
MIT-licensed open-source telemetry and memory layer for AI agents. It captures and normalizes prompts, tool calls, commands, file edits, approvals, reasoning, and usage across 20+ agent harnesses, and can export a portable record to existing observability and SIEM systems.
Managed runtime data platform built on Beacon with managed ingest, retention, search and analytics; fleet-wide inventory; identity, policy, approvals and governance; investigation timelines; real-time detection and response; and SSO, RBAC and audit trails.
Dedicated private-cloud or customer-controlled deployment with data isolation and residency options, custom retention and forwarding architecture, and coverage across endpoints, CI, and cloud environments.
Market Position
Asymptote positions itself as an open, cross-harness runtime data and control layer rather than a single-agent control plane. Its differentiation is the MIT-licensed Beacon telemetry foundation, a normalized portable activity record, broad endpoint/CI/cloud coverage, and an enterprise layer for runtime detection and blocking. The broader agent-security comparison set includes Straiker and Zenity, while existing SIEM/EDR and observability platforms are integration targets rather than being replaced.
Leadership
Founders
Justin D'Souza
Co-founder and CEO; previously led ML engineering at Level, Caffeine, and Doppel. At Doppel he built and scaled ML detection infrastructure from 1 million to more than 1 billion domains per day. He studied computer engineering at the University of Waterloo.
Shukan Shah
Co-founder and CTO; previously worked in payments at Level and Brex and in AI engineering at Doppel. At Doppel he led SOC automation and pioneered agentic AI for automated threat-hunting investigations. He studied computer science at Georgia Institute of Technology.
Executive Team
Justin D'Souza
Co-Founder & CEO
Former ML engineering leader at Level, Caffeine, and Doppel; built large-scale detection infrastructure at Doppel.
Shukan Shah
Co-Founder & CTO
Former payments engineer at Level and Brex and AI engineer at Doppel; led SOC automation and agentic threat-hunting infrastructure.
Founding Story
D'Souza and Shah had worked together at Level and Doppel and saw both the difficulty of building companies at the earliest stage and the infrastructure gap created by rapidly proliferating AI agents. Drawing on their experience in cybersecurity detection, SOC automation, and agentic threat hunting at Doppel, they started Asymptote to build a foundational runtime data layer that lets enterprises understand agent behavior, enforce controls, and improve performance. They joined the South Park Commons Founder Fellowship after Doppel.
Business Model
Revenue Model
Open-source Beacon is free and self-hosted. Asymptote monetizes managed enterprise runtime data, governance, detection and response, and private/customer-controlled deployments sold through custom enterprise contracts.
Pricing Tiers
Self-hosted Beacon with 20+ harnesses, portable telemetry, export, MDM rollout, and community support.
Managed ingest, retention, search and analytics; fleet-wide visibility; policy and approvals; investigation timelines; SSO, RBAC, audit trails; and priority support.
Enterprise capabilities plus private cloud or customer-controlled deployment, data isolation/residency options, and dedicated deployment support.
Target Markets
- Enterprise security and IT teams
- Engineering and developer-platform teams
- AI FinOps teams
- Regulated enterprises
- Organizations deploying coding agents on endpoints, in CI, and in cloud environments
- Security investigation and runtime protection for enterprise AI agents
- Engineering debugging, reliability improvement, and agent-performance optimization
- Platform-team inventory, adoption governance, and fleet rollout
- AI FinOps usage, efficiency, and token-cost analysis
- Regulated-enterprise governance of agents on endpoints, in CI, and in cloud environments
- Developer memory and self-improving coding workflows
- Security teams at regulated enterprises are reported as early production adopters; Asymptote has not publicly named those customers.