Fundraising Fox
Armur Ai
Techstars '23Bengaluru, IN · Founded 2023 · 5 employees on LinkedIn · 1 known investors
Find your way into Armur Ai
Sign up to see every warm intro you have to Armur Ai
- Paths you didn't know you had: your email and LinkedIn already hold routes to the Armur Ai team. We find them for you
- 2nd- and 3rd-degree connections: the friend-of-a-friend routes that take hours to manually find through your inbox or LinkedIn
- Answers now, not in days: momentum is everything in a raise. Skip asking around whether someone knows someone
Days of research, done the moment you sign in, with every route ranked by how warm it is.
Armur AI builds a swarm of autonomous AI agents designed to perform coordinated security testing and penetration testing on cloud infrastructure, APIs, and web applications. The platform chains vulnerabilities across multiple attack surfaces the way human attackers would, with reasoning that is auditable and reproducible.
Also known as Armur AI · Armur-Ai
Investors · 1
How we know: the Techstars portfolio · open dataset · Top 3000 Accelerator Startups (no YC) 2026-08 · Not right? Tell us
Company profile
researched Sep 2026Armur AI is an offensive-security company that builds AI-driven penetration testing infrastructure and runs security engagements against production systems. Its flagship project, pentest-swarm-ai, orchestrates eight specialist autonomous agents — recon, authentication, web, cloud, infrastructure-as-code, supply chain, network and exploitation — that share state through a stigmergic blackboard so that a finding from one agent becomes input for the next. The intent is to chain vulnerabilities across APIs, web applications and cloud infrastructure the way an adversary would, while keeping the reasoning reproducible and auditable; the project is written in Go, uses the Claude API and 7+ native security tools, is licensed AGPL-3.0, and supports bug bounty, continuous monitoring and CTF modes.
Around the swarm the company maintains a set of open-source modules: vibescan, a scanner for AI-generated ("vibe-coded") code that runs SAST, DAST and sandboxed exploit simulation across 15+ languages using 30+ tools including semgrep, nuclei, trivy, bandit, gitleaks and snyk; agat, a behavioral analysis pipeline for polymorphic and metamorphic malware; skillscan/claude-skill-scan, which audits Claude Skills for prompt injection, exfiltration paths and supply-chain issues; akira, a graph-based attack-path simulator mapping MITRE ATT&CK plus AI-specific vectors; and killswitch, an eBPF rootkit detector. The company also publishes the Pentest Dojo, a free, vendor-agnostic curriculum of eight labs (~14 hours, Apache-2.0, Docker Compose) covering agentic recon, autonomous exploitation, model-aware testing, prompt-injection chains, agent trust boundaries, swarm orchestration, AI supply chain and a live red-team capstone.
Earlier and adjacent positioning describes a developer security platform using LLM blending, agentic workflows and a mixture-of-agents approach for SAST, DAST and VAPT, with detailed plain-English vulnerability explanations, automatic fix generation and patching, IDE (VS Code) access and GitHub app/CI-CD integration, plus language- and stack-specific scanning (Go, Rust, Python, JavaScript, Solidity, Move, Solana, Dockerfile, Terraform, IaC) and blockchain security audits.
Founding story
Armur AI was founded in 2023 by Akhil Sharma; a Techstars listing gives a 2022 founding year and an initial Web3 cybersecurity focus.
Business model
Armur AI combines open-source offensive-security software with paid human-led services. The core tooling — pentest-swarm-ai, vibescan, agat, skillscan, akira, killswitch — is published on GitHub under AGPL-3.0, MIT and Apache-2.0 licences, while revenue-generating work comes from penetration-testing engagements in which the company's own operators deploy the swarm against a client's production environment, hand-validate every chained finding and deliver an audit-ready report. Earlier positioning also described a developer-facing code-security platform with a GitHub app, VS Code extension, pay-as-you-go pricing and team collaboration features.
Paid penetration-testing engagements (one-off or continuous cadence) delivered by the company's operators, alongside a developer security platform previously marketed with pay-as-you-go pricing; the open-source tooling itself is free.
Traction
The pentest-swarm-ai repository shows 2,624 stars and 480 forks and vibescan 84 stars and 15 forks as of 2026, across a GitHub organisation with 32 public repositories. A 2025 article reports the platform grew within 12 months to more than 4,000 engineers and 300 paying customers, with over 500 companies said to use it, naming Revolut, Airtable, Webflow and Vercel. Third-party profiles list 5 employees and about 6.16K monthly web visits.
Latest developments
Repository activity through 2026 shows continued work on pentest-swarm-ai (2,624 stars, AGPL-3.0), a Homebrew tap, killswitch (eBPF rootkit detector), claude-skill-scan and Akira. The team published XAMT: Adversarial Attacks on Multi-Agent Systems (arXiv:2512.15790) and presented agat at BSides Bangalore 2026, while advertising capacity for 10–15 hands-on engagements per year.
▸Full profile — market position, technology, go-to-market, geography, history, risks & controversies
Market position
Armur AI presents itself as an AI-native offensive-security vendor competing with established application-security scanners such as Snyk, Semgrep, SonarQube, Veracode, Checkmarx, Codacy, DeepSource, Embold and Aikido Security, differentiating on multi-agent LLM tooling and AI-generated-code coverage. Its GitHub organisation states backing from Techstars and Outlier Ventures.
The company positions the swarm's multi-agent, blackboard-coordinated design as able to chain findings across attack surfaces that single tools would not connect, and stresses that the code is open source so customers can audit the reasoning rather than trust a black-box scanner. Engagements are run by the same team that wrote the tooling, with hand-validated findings and reproducible run logs. On the code-scanning side, the company claims focus on vulnerabilities specific to AI-generated code, automatic fix generation and patching, and detailed explanations aimed at developers without a security background.
Technology
Multi-agent (swarm) architecture in which eight specialist agents coordinate via a shared stigmergic blackboard, using ReAct-style reasoning and the Claude API, implemented mainly in Go with 7+ native security tools; other modules are built in Rust and TypeScript. The broader platform is described as using LLM blending, agentic workflows and a mixture-of-agents approach across static analysis, dynamic testing and exploit simulation, with sandboxed exploit validation, SBOM/dependency and IaC (Terraform, Kubernetes) analysis, and automated fix generation.
Go-to-market
Distribution runs through open-source releases on GitHub (including a Homebrew tap for pentest-swarm-ai), self-serve sign-up and developer integrations such as a VS Code extension and GitHub app, a free training curriculum (Pentest Dojo), a Discord community and blog/tutorial content, and direct inbound briefs for paid penetration-testing engagements via akhil@armur.ai. Comparison pages target users of incumbent code-scanning tools.
Engineering, security and DevSecOps teams at companies running cloud infrastructure, APIs and web applications; developers shipping AI-generated code; and, per company statements, enterprises plus defense, intelligence and government organisations securing critical infrastructure.
Geography
Listings associate the company with San Francisco, California and with Bengaluru/Bangalore, Karnataka, India. Its agat malware-analysis tool was featured at BSides Bangalore 2026.
History
Company profiles record a 2023 founding by Akhil Sharma, with a pre-seed round dated 10 July 2023. A Techstars job-board listing describes an earlier Web3 cybersecurity product suite for decentralized applications and blockchain solutions using swarm learning. The product line subsequently centred on AI-driven penetration testing — first marketed as Pentest GPT alongside static and ML-based code scanning and blockchain security audits — and later on the open-source pentest-swarm-ai project and associated tooling, together with a services practice and the Pentest Dojo training curriculum.
Risks & controversies
Public profiles of the company are inconsistent: founding year is given as both 2022 and 2023, headquarters as San Francisco and as Bengaluru, and employee count as 5 or 6 across sources. Traction figures such as 4,000 engineers, 300 paying customers and 500 companies including named brands come from a single contributed-format article rather than audited disclosure, and the funding amount remains undisclosed. Services capacity is self-limited to roughly 10–15 engagements per year.
Compiled by commissioned research from 8 cited public sources — announcements, filings, and press listed under research sources below.
Key figures
latest reportedCompany-reported or press-reported figures, each dated to when it was claimed — not independently audited.
Related companies · 10
Companies working in the same space as Armur Ai.
Timeline · 5
launches, deals, and filingsagat, a behavioral analysis pipeline that fingerprints polymorphic and metamorphic malware, was featured at BSides Bangalore 2026.
Open-source autonomous penetration testing framework coordinating recon, classification, exploitation and reporting specialist agents through a stigmergic blackboard; written in Go with 7+ native security tools, licensed AGPL-3.0, and distributed via a Homebrew tap.
Open-source (MIT, Go) scanner running SAST, DAST and sandboxed exploit simulation across 15+ languages with 30+ tools, combining agents with semgrep, nuclei, trivy, bandit, gitleaks and snyk, plus AI-generated fixes and PR review integration.
Team publication on AI security and adversarial systems, posted to arXiv (arXiv:2512.15790).
Company profile lists a single funding round of an undisclosed amount, dated 10 July 2023, with the company at pre-seed stage.
Dated company events from announcements, filings, and press; legal rows summarize public dockets and regulator releases.
▸Research sources · 8
primary sources listed
- Armur Aiarmur.ai · web
8 public sources were cited for this profile; the first-party ones are listed here.
Frequently asked questions
- What does Armur Ai do?
- Armur AI builds open-source AI agent swarms for autonomous penetration testing and runs hands-on security engagements.
- Who are Armur Ai's investors?
- Armur Ai's investors include Techstars.
- Where is Armur Ai headquartered?
- Armur Ai is headquartered in Bengaluru, IN.