Security Research Lead

Summary of the Role: As Security Research Lead at Maze, your research directly shapes our products and how our AI understands real risk. This is a unique opportunity to join a well-funded Series A startup building at the intersection of generative AI and cybersecurity, leading the research function that makes our products sharper than anything else on the market — and gives Maze a credible technical voice in the security community. You'll set the bar for our vulnerability research: defining the methodologies that separate critical risk from noise, validating and contextualising the threats our AI surfaces, and feeding that expertise straight into product and engineering to improve how we detect, prioritise, and remediate. You'll work hand-in-hand with our AI/ML, product, and engineering teams — close to the roadmap and the code, not off to one side — turning what you find into capabilities that crush the competition. Alongside that, you'll amplify the work externally: surfacing novel vulnerabilities and building research narratives that earn real reach, giving Maze technical credibility with practitioners and customers. This role is perfect for a security researcher with real depth in cloud and application security who is ready to step up: someone who wants their research to ship as product, can lead a small research team, and can tell a story that lands with a technical audience. The ideal candidate has done this inside another security vendor and wants to build both great products and a research brand from a strong foundation. What you'll work on: - Make Our Products Brilliant: Feed research directly into product and engineering — work close to the roadmap and the codebase to sharpen how we detect, prioritise, and remediate, building capabilities that outclass the competition - Shape How Our AI Understands Risk: Translate deep threat research into the labels, signals, and product feedback that train our models to prioritise vulnerabilities like a seasoned researcher - Lead Our Security Research Function: Set the direction, standards, and methodologies for how Maze researches, validates, and prioritises cloud and application security threats, scaling a small team of researchers as we grow - Find Novel Vulnerabilities That Get Reach: Surface original research and build narratives — blog posts, technical talks, podcasts, video, conference presentations — that earn real reach and give Maze technical credibility with the security community - Build Authoritative Technical Intelligence: Produce detailed research on exploitation techniques, attack vectors, and remediation across cloud infrastructure and application security, enriched with CVE, advisory, and threat-intel sources - Set the Standard for Research Quality: Establish the frameworks and review processes that keep our vulnerability assessment consistent, defensible, and ahead of the threat landscape - Grow the Bench: Mentor and develop researchers, raising the technical bar of the team and creating a research culture others want to join What You Need to Be Successful: - Proven Security Research Depth:** 6+ years in hands-on security research, with a strong track record investigating complex vulnerabilities, building proof-of-concepts, and validating real-world threats - Cloud & Application Security Expertise: Deep knowledge of cloud (AWS) and application security vulnerabilities, attack vectors, and how they actually get exploited at scale - Research That Ships as Product: A track record of turning research into product and engineering outcomes — working closely with eng, product, and ML teams to translate findings into shipped capabilities, not reports that sit on a shelf - A Public Track Record: Demonstrated ability to be the voice of security research externally — published research, conference talks, well-received technical content, or a recognised presence in the community - Vendor-Grade Perspective: Direct experience inside a security tooling or research organisation, with a view on what good research operations look like and how to build credibility in market - Technical Investigation Skills: Strong coding and scripting (Python, Go, or similar) for automating research, building validation tooling, and creating PoCs - Communication That Lands: Ability to translate complex security research for technical practitioners, customers, and internal AI/ML and product teams without losing rigour - Leadership Instinct: Experience leading or mentoring researchers, setting standards, and owning a domain end-to-end in a fast-moving environment - Startup Readiness: Comfortable operating with ambiguity and limited structure, prioritising ruthlessly, and building the function as you run it - Nice to haves: - Experience with AI/ML security or working with AI-generated security findings - An established personal brand or following in the security research community - Open-source contributions to security tools or research - Experience setting up or scaling a research/labeling operation - Industry certifications (OSCP, AWS Security, CISSP, etc.) Why Join Us: - Ambitious Challenge: We're using generative AI (LLMs and agents) to solve some of the most pressing challenges in cloud and application security. You'll define how AI understands and prioritises vulnerabilities — and how the market talks about it. - Expert Team: We are a team of hands-on leaders with experience in Big Tech and Scale-ups, who have been part of the leadership teams behind multiple acquisitions and an IPO. - Build Products and a Research Brand: A rare chance to build a security research function from a strong foundation where your research ships as product — directly shaping what we build and how Maze is seen in the security community. - Impactful Work: Cybersecurity is a force for good. Your research will directly improve how thousands of organisations understand and respond to threats, scaling expert security knowledge through AI. - Build an AI-native Company: Join early enough to shape everything from the ground up, with significant equity upside and the room to grow into senior security leadership.