Manager, Cyber Forward Deployed EngineerKPMG LLP
Job Summary
At KPMG, the Client Technology & Innovation (CT&I) group transforms market disruption into scalable, agentic, software-powered businesses. We operate at the intersection of enterprise scale and startup velocity, shaping the firm's global product, platform, data, and AI agenda. Our teams focus on innovation at the edge: incubating bold concepts, co-innovating alongside frontier AI labs, and moving from idea to market at pace. Here, your work will not stay in demo mode. In this position, you can expect to design, build, and ship production software in rapid, real-world deployment cycles. If you are a hands-on creator who thrives in ambiguity and wants to build next-generation intelligent products, come build with us.
KPMG is currently seeking a Manager, Cyber ForwardDeployed Engineer to join our Clients, Technology & Innovation(CT&I) organization.
Responsibilities
- Own the technical vision and systemarchitecture for the pod's solutions, actively writing and shipping robustproduction code weekly while establishing high engineering standards forhardening, scalability, and live-environment reliability.
- Direct full-lifecycle offensiveengagements from scoping and strict legal rules of engagement through executionand executive readout, deploying frontier cyber models (e.g., Mythos/Fable,Google Argon, Daybreak) to uncover critical vulnerabilities, includingzero-days.
- Bridge the gap between finding flawsand fixing them by designing software updates, patches, and code refactors thateliminate root-cause technical debt and harden client applications.
- Develop and maintain the pod'sproprietary offensive tooling and agentic workflows, implementing rigorousevaluation harnesses, golden vulnerability datasets, automated regressiontesting, and human-in-the-loop safeguards.
- Drive deliverables in fast-pacedtwo-to-six-week delivery cycles, actively managing risk, identifying technicalconstraints early, and ensuring every release is production-ready and fullycontained within authorized parameters.
- Lead technical and executive workingsessions across engineering groups, CISOs, and C-suite stakeholders, whilecoaching pod engineers to sustain talent density, quality, and execution speed.
- Integrate automated security testing,policy-as-code, and compliance gates directly into the Harness engineeringpipeline to catch vulnerabilities before deployment
- Build specific defense profiles fordual-model setups to include mitigating prompt injection and data poisoning forLLMs, while ensuring input validation, algorithmic tampering defenses, andbypass prevention for deterministic Type 1 models (e.g., Jev)
- Analyze customer journey path modelsto discover logical vulnerabilities, privilege escalation vectors, or stateswhere non-deterministic LLM behavior can break the intended business logic flow
- Act with integrity, professionalism,and personal responsibility to uphold KPMG's respectful and courteous workenvironment.
Qualifications
- Minimum seven years of recent,hands-on offensive security experience (red teaming, penetration testing,exploit development, or vulnerability research) paired with active, currentproduction software engineering capability (e.g., Python, Go, Rust, or Java);published research, CVEs, or certifications (OSCP, OSEP, OSCE3, GXPN) arestrong pluses
- Minimum 3+ years of hands-onleadership managing, mentoring, and scaling high-performing engineering teams;thrives with ambiguity and a bias toward shipping outcomes (founder orearly-stage startup experience is a strong plus)
- Bachelor's degree in a STEM field oreight years equivalent practical experience
- Demonstrated experience building,hardening, and operating scalable systems in live enterprise environmentsacross Azure, AWS, or Google Cloud (GCP), with proven ability to refactorlegacy code and remediate production defects
- Hands-on fluency with frontier AImodels, modern coding environments (e.g., Claude Code, Cursor), and agenticsystems, including context engineering, tool-calling frameworks, evaluationdiscipline, and protocols such as MCP (Model Context Protocol); directexperience engineering agent harnesses or contributing to open-source developerinfrastructure is highly valued
- Demonstrated ability to articulatecomplex security findings and AI technical concepts to both hands-on engineersand C-level executives (CIO, CISO, CFO), translating security and operationalinsights into quantified business value
- Impeccable ethical judgment regardingrules of engagement, authorization, and sensitive data handling
- Accessible portfolio, GitHubrepository, or system walkthrough demonstrating shipped work
- Ability to travel domestically asneeded