Trust Architect

Job description

SUMMARY Trust Architects build trust in the AI and trust among the people who use it. They make sure every system Avanade Next ships is safe, reliable and ready for production — and that the client's people are confident, capable and willing to adopt new ways of working alongside it. The role operates across two dimensions. Trust in the AI: evaluation frameworks, accuracy measures, security controls, governance controls and human-in-the-loop safeguards. Trust in the people: adoption design, capability uplift, ways of working and confidence in human-agent collaboration. KEY RESPONSIBILITIES Building trust in the AI Design and run evaluations: define accuracy targets, build the harness, drive the system to target and evidence it. Set governance, risk and security guardrails before build begins, and keep them current as the system evolves. Embed guardrails, controls and human-in-the-loop review, including the feedback loop that improves the system after release. Establish the evidence required for a system to move from prototype to production, and hold the release gate. Own the regulatory and privacy position on the engagement — APRA, MAS, data residency and client-specific obligations. Work shoulder to shoulder with FDEs so trust, evaluation and monitoring capability is built in from the outset. Building trust in the people Design the adoption and change approach that builds confidence through transparency, education and demonstrated outcomes. Help the client redesign ways of working and human-agent interaction so the system fits how people actually work. Lead enablement so the client's people can and do use what has been built. Drive uptake and user confidence after go-live, tracking active usage rather than reported satisfaction. Support executive and Board confidence with clear governance, accountability and decision-making frameworks. Contribute evaluation harnesses, guardrail patterns, governance frameworks and adoption assets to the APAC library. KEY MEASURES Governance, guardrails and security controls in place before build starts on every engagement. Target accuracy, safety and governance thresholds met and evidenced before production release. Adoption above 70% active usage at 90 days post-go-live, with effective human-in-the-loop operation. Full compliance with Avanade responsible AI standards and client regulatory obligations. Reusable trust, evaluation and adoption assets contributed from every engagement. . SKILLS AND EXPERIENCE 5+ years across AI governance, model risk, security, responsible AI, or large-scale technology change and adoption. Strength in two of these with a credible path into the third is expected — few candidates will have all of it. Hands-on evaluation experience with generative or agentic systems is strongly preferred over policy-only backgrounds. Working knowledge of the APAC regulatory environment relevant to their market. Practical change and enablement capability with technical audiences and frontline users alike. Able to build trust with sceptical stakeholders and to say no with evidence. Comfortable in a role that is still being defined, in a track with no established internal precedent. WAYS OF WORKING Engaged from the earliest stage of the engagement, not the week before go-live. Balances trust in the AI and trust in the people who use it — evaluation and governance alongside change and adoption. Held centrally in the shared workforce pool and deployed across all of Next. Works shoulder to shoulder with FDEs during build — safety is designed in, not inspected in. Provides independent challenge where trust, security, governance or adoption risks are not being adequately addressed, while helping the team find a practical path forward. Stays with the client through adoption, so responsibility does not end at deployment. AI-native by default in their own work; regional travel following the mandate. .