Tolmo (tolmo.com): The AI Security Agent Fleet | $22M Seed
AI coding agents are dramatically accelerating how fast software is written, reviewed, and shipped into production. Traditional application security tools and processes — built for slower, human-paced development — are increasingly unable to keep up. Attackers are already exploiting vulnerabilities in an average of 1.5 days, while many security teams still operate on timelines measured in weeks.
Tolmo is building the missing layer: autonomous AI security agents that operate on a live, connected knowledge graph of your entire production stack.
As a company that just came out of stealth in mid-June 2026 with a substantial $22M seed round, Tolmo is a prime example of why human-curated editorial profiles outperform generic data aggregators. This piece delivers timely context, founder pedigree analysis, competitive positioning, and proprietary insight that automated scrapers cannot replicate.
Data
Funding Stage: $22M Seed round led by Accel, with participation from Y Combinator. Announced in June 2026 as the company exited stealth. This is a sizable seed for the category, reflecting strong conviction in the thesis and team.
Launch / Founding Date: Founded 2026 (YC Spring 2026 batch). Public launch and major funding announcement in June 2026.
Key Leadership:
- Pierre Betouin, Co-Founder & CEO — Previously led security products at Datadog following the acquisition of Sqreen (YC W18), a cloud and web app security monitoring platform that protected over 800 customers in production.
- Vladimir (Vlad) de Turckheim, Co-Founder — Staff Engineer on security products at Datadog. Founder of Heal.dev.
- Jean-Baptiste (JB) Aviat, Co-Founder
- Arnaud Breton, Co-Founder
The early team includes strong technical talent with deep production security and observability experience. Team size is approximately 12, based in San Francisco.
Core Tech Stack / Approach: Agent-driven security platform built around a live knowledge graph that continuously connects code, cloud infrastructure, CI/CD pipelines, observability data, and security vendor signals — while maintaining a full history of changes (“time machine for production environments”). A fleet of specialized autonomous AI security agents (including pentester, cloud config, and security monitor agents) operates on this graph. The system ingests data via read-only integrations with no agents to deploy and no code changes required. Agents run on pull requests, deployments, and alerts to provide contextual findings rather than isolated signals.
Editorial
Plain English Pitch (2 sentences):
Tolmo acts like a team of specialized AI security experts that never sleeps and has perfect, up-to-the-minute knowledge of your entire production environment. It automatically builds a connected “map” of your code, cloud setup, pipelines, and monitoring tools, then deploys smart agents that watch every change, spot real risks in context, and can even help respond to attacks 24/7 — without requiring you to install anything or rewrite code.
ICP & Primary Use Cases:
Primary buyers are security, platform, and engineering teams at companies that are aggressively using AI coding agents (or simply shipping code very quickly). These organizations face a widening gap between development velocity and their ability to secure what gets deployed.
The core problem Tolmo solves is the loss of visibility and control as AI-generated code changes flood into production faster than traditional AppSec processes or tooling can handle. It delivers production-context-aware detection, prioritization, and response at machine speed.
Key use cases include securing AI-accelerated development pipelines, contextual vulnerability management across the full stack, automated detection of risky configurations or changes, and rapid incident response with full historical and environmental context.
Hiring Patterns:
Post-$22M seed and stealth exit, Tolmo is in active growth mode with a team of ~12. Expect focused hiring in AI/security engineering, agent development, product, and go-to-market roles as they expand beyond the initial 45 design partners. This pattern signals rapid scaling of both the product and commercial efforts.
Buying Signals:
- Major $22M seed announcement and stealth exit in June 2026.
- Backing from top-tier investors (Accel + YC).
- Strong founder pedigree from Datadog security (post-Sqreen acquisition).
- Early traction with 45 design partners iterating on the platform.
- Clear positioning around the urgent, timely problem of AI-driven development outpacing security.
Proprietary Insights
Proprietary Score — AI-Native Production Security Readiness Index:
Tolmo scores exceptionally high on this custom early-stage metric. Key drivers include the founders’ deep, relevant experience securing production systems at Datadog scale, the extremely timely thesis (AI code generation creating new attack surfaces and velocity challenges), high-quality investor validation, and meaningful early design partner engagement. The combination of strong technical DNA in observability + security with a clear “agent fleet on a live production graph” architecture positions the company well in a rapidly evolving category.
Competitor Matrix (Editorial Comparison):
| Dimension | Tolmo (Agent Fleet + Live Knowledge Graph) | Traditional AppSec (Snyk, Veracode, etc.) | Cloud Security Posture / CNAPP (Prisma, Wiz, etc.) | Observability + Security (Datadog Security) | Manual / Human-Led Security Reviews |
|---|---|---|---|---|---|
| Core Strength | Contextual, agentic automation across the full production stack at AI speed | Strong static/code analysis | Strong cloud configuration & posture | Excellent visibility & correlation | Deep expertise but slow |
| Production Context | Very High (live connected graph + history) | Low to Medium | Medium | High | High (but limited scale) |
| Speed vs. AI Development | Designed for AI-accelerated velocity | Not optimized for AI code gen pace | Moderate | Moderate | Too slow |
| Automation Level | High (autonomous agents + 24/7 team) | Medium | High (posture) | Medium | Low |
| Current Stage | Just out of stealth, $22M seed | Mature | Mature | Mature | Evergreen |
| Best For | Teams shipping fast with AI coding tools | Code-level vulnerability management | Cloud infrastructure security | Visibility & investigation | High-trust, low-volume environments |
Founder & Company Vision Highlights (Public sources only):
The core thesis is that security must evolve to match the new reality of AI-driven software development. Traditional approaches cannot keep pace with how quickly code is now generated and deployed, creating dangerous blind spots in production. Tolmo’s approach — a live, connected knowledge graph plus a fleet of specialized security agents — aims to give security teams the same (or better) speed and context advantages that attackers and AI coding tools currently enjoy.
Deeper proprietary perspectives on specific agent capabilities, integration depth with popular AI coding tools, response workflows, and roadmap priorities are best obtained through direct conversations with the founding team.
Why This Matters Now
The rapid adoption of AI coding agents is one of the most significant shifts in software development in years. It supercharges velocity but simultaneously compresses the time security teams have to understand and mitigate risk. Tolmo is one of the first platforms purpose-built for this new operating environment — combining deep production context with autonomous agents that can operate at machine speed.
