About AppViewX:
AppViewX is the only machine and agent identity security company built for the AI and quantum era, bringing together discovery, automation, control, and intelligence. The AVX Platform helps enterprises reduce risk by providing complete visibility and governance over every machine and AI agent identity, automating their lifecycle, controlling their access, and proving compliance at the speed business demands. Trusted by global enterprises across financial services, healthcare, and technology, AppViewX is recognized as a leader in the IDC 2026 MarketScape for Certificate Lifecycle Management and KuppingerCole’s 2025 Non-Human Identity Management Leadership Compass. For more information, users can visit appviewx.com.
Our Values
Our values define how we work, make decisions, and deliver for our customers and each other. Some reflect strengths deeply engrained in how we work. Others represent the standards we are committed to building together.
- We Do What We Say: We own our commitments, communicate honestly about progress, and measure success by results, not activity.
- We Are Trusted Experts: We continuously build our expertise, offer candid guidance, and focus on the outcomes our customers and colleagues need.
- We Raise the Bar: We stay curious, challenge the status quo, take smart risks, and turn better ideas into meaningful results.
- We Win Together: We share information, debate openly, commit together, and recognize that our success is connected to the success of our customers and teammates.
If these values reflect how you want to work and contribute, you will find an opportunity to make a meaningful impact at AppViewX.
Role Overview:
AppViewX is building Agent Identity Security and shipping AI-enabled features into products that secure the machine and agent identities of the world’s largest enterprises. Internally, every business unit is moving to AI-augmented and increasingly agentic workflows. Our customers, auditors, and board expect us to be the most credible practitioner of AI security in our category, not just a vendor of it.
This role has two mandates that reinforce each other. First, secure AI and agentic AI: the models, agents, tools, MCP servers, data pipelines, and identities across both our products and our corporate environment. Second, use AI and agents to secure AppViewX: automate detection, triage, testing, and assurance so a lean team operates at a scale that would otherwise take several times the headcount. This is a Senior-level role: you are strong across every domain of security and privacy engineering, you operate with minimal direction in ambiguity, and you mature the technical and administrative controls of the whole organization as the technology changes beneath us. You are an engineer who has personally broken and hardened agentic systems in production, and who has personally built agents that do real security work.
Key Responsibilities:
Securing AI and agentic AI
- AI security threat modeling. Author and maintain threat models for AI-enabled product changes and internal agent deployments, covering prompt injection (direct and indirect), tool misuse, excessive agency, data exfiltration, memory and context poisoning, and cross-agent trust. Align to OWASP LLM Top 10, OWASP Agentic AI threats, and MITRE ATLAS.
- Agent identity, permissioning, and tool-call controls. Define and implement least-privilege, short-lived, auditable identities and scoped credentials for agents; human-in-the-loop gates for high-impact actions; and sandboxing, egress control, and policy enforcement around tool calls and MCP servers. This is our core domain, and we should be the reference implementation.
- Prompt and output safety in product. Build input and output guardrails, content and data-leak filters, and structured-output validation into AI-enabled product features, delivered as reusable components engineering can adopt by default.
- AI penetration testing and adversarial testing. Lead hands-on penetration tests and adversarial campaigns against AppViewX AI features, internal agents, RAG pipelines, and MCP servers, covering every category of the OWASP Top 10 for LLM Applications and the OWASP agentic threat taxonomy: prompt injection, insecure output handling, training data and context poisoning, model denial of service, supply chain, sensitive information disclosure, insecure plugin and tool design, excessive agency, overreliance, and model theft. Use manual tradecraft plus automated attack agents and fuzzing, scope and run external AI pen test vendors, and turn every finding into a fixed control, regression test, and detection, not a slide deck.
- Product and corporate scope. Cover both sides of the house: AI security for AppViewX products (including Agent Identity Security) and for the corporate environment (employee AI tools, agents, connectors, and the data they touch). Feed AI security findings into Jira and the SOC alert pipeline so product and corporate risk is managed in one view.
- AI supply chain security. Partner with engineering on model, dataset, and dependency provenance: model and weights integrity, third-party model and agent-platform risk, MCP server and plugin vetting, and AI-BOM practices.
- AI security in the SDLC. Embed evals, safety regression tests, and policy checks in CI/CD so secure AI patterns are the fastest path for engineers.
Using AI and agents to secure AppViewX
- Security agents in production. Design, build, and run agents for vulnerability triage and prioritization, exploitability and reachability analysis, SAST/SCA finding validation, alert enrichment, and incident evidence gathering. Every agent has a defined scope, measured precision and recall, a human escalation path, and an audit trail.
- AI-assisted detection and response. Integrate AI security telemetry into the detection pipeline with the SOC team, author detections for AI and agentic attack patterns, and help the MDR and AISOC tooling absorb more L1/L2 work safely.
- AI-native AppSec. Work with Product Security on AI SAST, AI-assisted code review, and automated test generation for product security, with false-positive and developer-friction metrics owned and published.
- GRC and customer-trust automation. Partner with the GRC team to automate evidence collection, control testing, and questionnaire response with agents, with accuracy measured and tuned.
- Safe internal AI enablement. Help assess and approve AI tools, connectors, and agents used by employees (including Claude and MCP integrations): scoped permissions, data boundaries, logging, and monitoring that let the business adopt quickly without creating shadow AI risk.
- Leverage measurement. Report AI leverage ratios, such as work absorbed by agents, mean time to triage and remediate, and agent accuracy, so leadership can see the return and the residual risk.
Governance and customer trust
- Control evidence. Map AI security controls to NIST AI RMF, ISO/IEC 42001, and AIUC-1 so our AI security posture is audit-ready and defensible in customer reviews.
- Customer engagement. Support the AI Security Lead and Sales Engineering in AI security deep-dives, trust reviews, and RFP responses with credible technical depth.
What You Bring/Qualifications:
- Full-stack security and privacy engineering. Act as the technical lead across AppSec and product security, cloud and infrastructure, identity and secrets, vulnerability and exposure management, detection and response, and privacy engineering (data minimization, DPIAs, retention, data subject requests, sub-processor and DPA controls under GDPR). Go deep where the risk is, and know when to pull in specialists.
- Technical and administrative controls. Own the maturity path for both: engineering controls (guardrails, pipelines, configurations, monitoring) and administrative controls (policies, standards, procedures, exception handling, training, and ownership). Make them measurable, mapped to SOC 2, ISO 27001, NIST CSF 2.0, and PCI DSS, and operating continuously rather than at audit time.
- Ambiguity and change. Take loosely defined problems in a fast-moving AI landscape, frame the risk, propose options with trade-offs (accept, mitigate, defer), make a recommendation, and drive it to done. Revisit decisions as models, agent frameworks, regulations, and attacker tradecraft evolve.
- Technical leadership without hierarchy. Set engineering standards, review designs, coach AppSec, ProdSec, and Red Team engineers, and mentor junior engineers in India. Influence Product, Engineering, SRE, and IT leaders through credibility and clear business-impact reasoning.
- Incident and surge readiness. Serve as a technical lead in high-severity incidents and vulnerability surges, including AI-related events, and turn post-incident learning into durable control improvements.
Required
- Six or more years in security engineering, including senior or equivalent scope, and at least two years of hands-on, production-level work securing LLM and agentic AI systems (not just policy or assessment).
- Strength across the breadth of security and privacy engineering: AppSec, cloud, identity, vulnerability management, detection and response, and privacy engineering, with deep expertise in at least two.
- Track record of maturing both technical and administrative controls across an organization (policies, standards, guardrails, evidence), in a high-growth SaaS environment.
- Proven ability to operate in ambiguity: you set direction without a playbook, make defensible risk-based calls, and keep delivering as the landscape shifts.
- Demonstrated depth in the LLM and agent stack: prompt injection and jailbreak classes, tool-calling and function-calling risks, RAG and vector store security, memory and context handling, MCP and agent protocols, and multi-agent trust boundaries.
- Expert-level AI penetration testing and adversarial testing: you have attacked real LLM applications and agentic systems end to end, found and chained exploitable issues (for example indirect prompt injection to tool abuse to data exfiltration), written clear reproducible reports, and driven remediation with engineering.
- Mastery of the OWASP Top 10 for LLM Applications, with practical attack and defense experience across every category, plus OWASP agentic AI guidance and MITRE ATLAS.
- Fluency with AI red-team tooling and methods, such as Garak, PyRIT, Promptfoo, or equivalents, and the ability to build custom attack harnesses and evaluation suites.
- You have built and operated AI agents or LLM-powered tooling that perform security work, with evaluation and guardrails, and can speak to what failed in production and what you changed.
- Strong Python for security tooling, automation, and adversarial testing; comfortable with APIs, CI/CD, and cloud-native environments.
- Hands-on with at least one AI security platform or LLM evaluation and guardrail framework.
- Working knowledge of NIST AI RMF and the ability to translate OWASP, MITRE ATLAS, and NIST guidance into engineering controls.
- Sound judgment on identity and least privilege for non-human actors: OAuth, workload identity, secrets handling, and scoped, short-lived credentials.
- Self-starter and go-getter: you find the problem, own it, and finish it, without waiting to be assigned.
- Clear written and verbal communication; you can brief a CISO, challenge an engineering lead, and write a threat model another engineer can implement.
Preferred
- Hands-on agentic AI development, or published adversarial AI research, AI/ML CTF results, or AI bug bounty and responsible disclosure credits.
- Contributions to OWASP, MITRE ATLAS, open-source AI security tools, or public vulnerability disclosures in AI systems.
- Familiarity with ISO/IEC 42001, AIUC-1, or similar AI assurance frameworks.
- Background in PKI, certificate lifecycle, or machine identity (directly relevant to our product domain).
- Experience in detection engineering or SOC automation with SIEM, SOAR, or AI SOC platforms.
- Hands-on experience with SOC 2, ISO 27001, PCI DSS, or GDPR programs in a regulated B2B SaaS environment; privacy certifications (CIPP, CIPT) or security certifications (CISSP, OSCP, CCSP) a plus.
- Prior work across US and India teams.
Why AppViewX?
AppViewX caters to a wide range of customers from Fortune 1000 companies, including six of the top ten global commercial banks, five of the top ten global media companies, and five of the top ten managed healthcare providers. Over the years, we grew our diverse team, perfected our automation platform, and expanded our global footprint to India, North America, United Kingdom, and Australia. Today, we are headquartered in New York City and have come a long way by optimizing opportunities to create lasting relationships with enterprises, gaining unshakable customer trust along the way.
AppViewX is proud to be an Equal Employment Opportunity Employer. It is AppViewX’s policy to afford equal employment opportunities to all employees regardless of race, color, national origin, ancestry, religion, citizenship status, gender, gender expression or identity, sexual orientation, age, marital status, military or veteran status, pregnancy, disability, genetic information, arrest record, or other protected class under state, federal, or local law.
