Description
GovTech is seeking a strategic and mission-driven Lead/Senior Product Manager to spearhead the AI Safety and Governance pillar within PlatformAI. As AI adoption scales across the public sector, your mission is to ensure that every AI-enabled application is secure, fair, and accountable by design.
You will own a dual-track roadmap:
- For Developers: Building the "Safety Stack",a suite of automated Evals, Guardrails, and Red-teaming tools that allow teams to deploy LLMs with speed and confidence.
- For Governance and Observability: Architecting the "Governance Layer", observability tools and risk-posture dashboards that provide a unified view of AI compliance across the whole-of-government (WOG) ecosystem.
Key Responsibilities:
- Defining the Sovereign Safety Standard
- Define the long-term strategy for AI Safety-as-a-Service, transitioning governance from a manual checklist to an automated platform feature.
- Synthesize emerging global frameworks and local AI governance policies into platform features that make compliance "invisible" for developers.
- Establish the government’s standard for AI risk levels, determining the safety benchmarks required for different use cases (e.g., citizen-facing vs. internal productivity).
- Building the Guardrails & Audit Trails
- Lead the development of automated evaluation engines that stress-test products and models for prompt injection, PII leakage, toxicity, and hallucination before and after deployment.
- Partner with Engineering to integrate safety checks directly into CI/CD pipelines, ensuring that "Safety-by-Design" accelerates rather than hinders deployment.
- Build real-time monitoring tools to detect "jailbreak" attempts and model drift, ensuring rapid mitigation of live AI risks.
- Bridging Policy & Code
- Act as the primary liaison between Policy/Legal teams and Engineering teams to ensure technical tools meet policy and legal intent.
- Cultivate a culture of safety across GovTech by publishing playbooks, standardizing red-teaming exercises, and educating teams on the evolving LLM vulnerability landscape.
- Trust-Centric Metrics & Accountability
- Define and track KPIs such as Mean Time to Detect (MTTD) vulnerabilities, Policy Coverage %, and Guardrail Efficacy Rates.
Requirements:
- Senior Product Leadership: Proven experience managing complex technical products, specifically those focused on Security, Governance, or AI/ML infrastructure.
- AI Safety Domain Expertise: Deep understanding of Generative AI risks (e.g., prompt injection, training data poisoning, algorithmic bias) and the current state of AI safety research.
- The "Translator" Skillset: Exceptional ability to distil abstract ethical principles into concrete technical requirements and APIs.
- Analytical Rigor: A background in using data-driven evaluation frameworks (e.g., RAG evaluation, perplexity scoring, or model benchmarking) to measure performance and risk.
- Developer Empathy: A deep understanding of developer workflows and a commitment to making safety tools an "accelerant" rather than a "roadblock."
Benefits:
GovTech offers a total rewards approach, providing leave benefits, employee wellness programs, and flexible work arrangements.