Director, Enterprise Systems & AI Architecture

Bureau Veritas•Houston, TX

5d•Onsite

About The Position

Bureau Veritas is building an operating system for trust: the platform that scales expert judgement across the company and makes our findings more consistent and more defensible. You will decide how it fits together — data platforms, APIs and integration, cloud infrastructure and AI services — and you will sign off the interfaces between systems owned by different teams and into our shared AI backbone. Bureau Veritas runs AI in production and is building more of it in more places every quarter, across six very different businesses: Marine & Offshore, Industry, Buildings & Infrastructure, Certification, Consumer Products, and Agri-Food & Commodities. We are an accredited certification body, and our findings carry legal and commercial weight, so auditability and traceability are engineering constraints here. This is hands-on work designing and building the systems, rather than just oversight.

Requirements

12+ years of experience in enterprise architecture, software engineering, AI platforms, data platforms, or digital transformation.
Strong expertise in cloud architecture, APIs, integration patterns, data platforms, and AI/ML technologies.
Experience designing and implementing enterprise-scale technology solutions.
Proven ability to influence senior stakeholders and lead cross-functional initiatives.
Excellent communication, leadership, and strategic planning skills.
Real depth in all four areas of data, APIs and integration, infrastructure, and AI.
Past API and event-driven integration designs between systems owned by different teams; data contracts, and how data platforms behave under load; cloud infrastructure, and how hosting and identity choices constrain everything built on them; AI serving and evaluation, and how a model's output affects the rest of a system.
Past architecture you designed that other people built, deployed to production at scale and that held.
Experience explaining architectural design decisions to a business-line leader in terms of their own results.
Experience with hands-on AI work: retrieval and grounding, evaluation, agent and tool boundaries, and the judgement to say when a foundation model is the wrong answer.
Battle-tested standards you set for teams outside your reporting line, with their own budgets, and the methods you used to make them hold.
AI tools or agents as part of how you and your team work day to day, and a clear view of what you check rather than trust.

Nice To Haves

Architecture that other teams chose to adopt when they did not have to.
Systems that work where connectivity does not: intermittent, offline-first or edge-constrained delivery, and the data-consistency decisions that come with it.
Senior technical hiring depth, with a bar you can describe.
Regulated, safety-critical or audited delivery, where traceability shaped your engineering decisions.

Responsibilities

Own how data, APIs, infrastructure and AI fit together across our AI builds, and how they integrate with the applications our other departments run, and sign off the interfaces between separately owned pieces.
Build the shared backbone that delivery teams inherit: orchestration, retrieval and grounding, tool surfaces, evaluation harnesses and AI serving.
Decide where an AI output may feed a certified decision, where it may assist an accountable person, and where it must not be used at all.
Set the standard for what gets built with what: when the right answer is a deterministic rule, a small model we run ourselves, or a large foundation model. These standards need to be clear enough to be followed across all teams without you in the room.
Hire and lead a team of solution and database architects, and run design reviews.
Work directly with, and build consensus among, the leaders of our business and product lines, who determine what reaches production for their area.
Work as a peer with the directors of our AI engineering sites and with our technology, data science, information security and AI product leaders.
Build on AWS, with model access through Amazon Bedrock and infrastructure as code in Terraform.