Senior Director, AI Architect

DISA Global Solutions is an industry-leading safety and compliance solutions provider with customers across the Globe since 1986 and more than 1,200+ Team members across more than 30 locations. When you join the DISA Team, you join an industry leader that more than 30% of fortune 500 companies use. With a rich history of IT innovation, we have more than 55,000 customers and complete more than 10 million orders each year. 

DISA is proud to be a “Top Workplaces” award winner across multiple locations in 2023. We offer competitive pay, a robust benefits package, and an exciting place to work. We have a collaborative and team-oriented culture with numerous opportunities for career advancement. We host monthly appreciation events, and our internal learning and development team provides the resources for employees to engage in continuing education and training. 

DISA operates at the center of employment background screening, occupational health, and drug and alcohol testing — processes that are document-heavy, regulated end to end, and executed at very high volume. The Senior Director, Artificial Intelligence, owns how AI is applied across that operating footprint: what gets built, on what enterprise foundation, under what controls, and with what measurable business return.

This is a leadership role. You will identify and propose AI use cases, carry them through Technology, Product, Security, and Compliance to get the approvals that make them real, and then see them delivered — and you will be in the architecture reviews, the model evaluations, and the code yourself.

You will lead an in-house AI team that grows as the portfolio does, and you will use every delivery route available: your own team, existing engineering capacity elsewhere in the organization, vendor partners, or a ready-made product where buying beats building. You own the recommendation and the outcome regardless of which route it takes.

• Define and maintain the enterprise AI strategy and multi-year roadmap, tied to measurable outcomes: turnaround time, adjudication accuracy, analyst productivity, and client retention.

• Serve as the primary AI advisor to business and product stakeholders — translating between technical possibility and commercial consequence in both directions, and helping teams see where AI genuinely changes the economics of their process and where it does not.

• Drive AI acceleration across the organization: build awareness of what is now possible, create pull for AI adoption in the business, and move teams from curiosity to production use.

• Run structured build-vs-buy-vs-partner evaluations for AI platforms and vendors; own the recommendation, the total cost model, and the exit path.

• Own AI portfolio investment: prioritization, budget, FinOps discipline on inference and compute spend, and regular value reporting to leadership.

• Architect and deliver the enterprise AI framework — the shared foundation of model access, retrieval, orchestration, evaluation, observability, prompt and context management, and reusable agent patterns that product teams consume rather than rebuild.

• Lead the multi-agent orchestration platform and its supporting services from prototype maturity to production-grade, multi-tenant, environment-promoted (dev/staging/prod) infrastructure.

• Establish LLMOps and MLOps practice: CI/CD for models and prompts, regression and eval harnesses, drift and hallucination monitoring, human-in-the-loop feedback capture, incident response, and rollback.

• Define reference architectures, integration standards, and API contracts; represent AI work at the Architecture Review Board.

• Set the technology direction across a dual-platform foundation — Azure AI Foundry / Azure OpenAI and AWS Bedrock — treating both as first-class options and favoring managed services over bespoke development where the trade-off holds.

• Architect for model and platform portability: a common abstraction for model access, retrieval, and orchestration so workloads can move between Azure and Bedrock on capability, latency, cost, data residency, or contractual grounds without a rewrite.

• Make the platform call per workload on evidence — benchmark model performance and unit economics across Azure AI Foundry and Bedrock's model catalog (Anthropic Claude, Amazon Nova, Meta Llama, Mistral and others) rather than defaulting to incumbency.

• Bring depth in the surrounding services on both clouds: Azure AI Search, Document Intelligence, Azure Machine Learning, Fabric; Amazon Bedrock Knowledge Bases and Agents, Textract, SageMaker, OpenSearch.

• Background screening: intelligent document extraction from court and agency records, criminal record parsing and normalization, identity resolution and matching, verification automation (employment, education, licensure), adjudication decision support, and adverse action workflow assistance.

• Drug and alcohol testing: chain-of-custody document processing, lab result ingestion and reconciliation, MRO workflow assistance, random selection pool intelligence, collection site network optimization, and exception detection.

• Occupational health: clinical document understanding, clearance and compliance status inference, and scheduling optimization.

• Cross-cutting: conversational and agentic interfaces over enterprise data, contact center deflection and agent assist, internal engineering productivity tooling, and analytics/metadata intelligence across the reporting estate.

• Own the enterprise AI governance framework: model inventory and registry, risk tiering, pre-deployment review, bias and fairness testing, explainability requirements, human oversight thresholds, and audit evidence.

• Ensure AI systems operate within the regulatory perimeter DISA works in — FCRA and state screening law, DOT testing regulations, HIPAA, SOC 2, and GDPR — in partnership with Legal, Compliance, Privacy, and Security. Regulatory interpretation stays with those functions; your job is to make the controls real in the architecture.

• Establish data governance for AI: lineage, minimization, retention, tenant isolation, and controls on the use of consumer and candidate data in training and retrieval.

• Source, shape, and prioritize AI use cases across the business; build the business case for each — problem, expected value, cost, risk, and success criteria.

• Take proposals through the stakeholder path to approval: Technology for architecture fit, Product for roadmap alignment, Security for control adequacy, Compliance for regulatory exposure. Bring these functions in early rather than presenting to them late.

• Choose the delivery route on merit for each initiative — build with your own team, build with engineering capacity elsewhere in the organization, build with a vendor or systems integrator, or buy a proven product — based on time to value, total cost, differentiation, and long-term supportability.

• Own vendor selection and management end to end: evaluation criteria, proof of concept design, commercial and security review, contract input, and ongoing performance against agreed outcomes.

• Stay hands-on through delivery: prototype to prove feasibility before committing budget, review designs and code, and validate model behavior yourself rather than accepting it on report.

• Build and lead the AI team at the Mumbai GCC — AI/ML engineers, data scientists, and AI platform engineers — using third-party and contract hiring to stand up capacity quickly rather than waiting on full-time requisition cycles.

• Work closely with staffing partners on sourcing, technical screening, and onboarding; set the technical bar and hold it, so speed of hiring never comes at the cost of engineering quality.

• Blend contract and full-time capacity deliberately: use third-party resourcing to move fast and absorb variable demand, and convert or backfill with permanent hires where the capability needs to stay in-house.

• Partner with vendors and systems integrators as an extension of the team — set the architecture and standards they build to, embed them in your delivery rhythm, and manage them to outcomes rather than hours.

• Ensure knowledge stays with DISA when contractors and vendors roll off: documentation, code ownership, and handover discipline built into every engagement from the start.

• Establish the AI Center of Excellence — enablement, internal standards, reusable assets, and AI literacy programs for engineering and business teams.

• Coach and grow technical leaders; set the bar for engineering rigor and for saying no to AI that does not earn its place.

Required

• 15+ years in technology, with 6+ years leading AI/ML or data science organizations, including leadership of leaders.

• Proven record of implementing AI solutions at enterprise scale within a product organization — AI embedded in commercial products serving external customers, not internal experiments or consulting engagements. You understand product roadmaps, release cycles, multi-tenancy, SLAs, and what it means to support an AI feature customers pay for.

• Demonstrated delivery of AI systems in production, not pilots. You can describe what you shipped, what it cost, what it returned, and what broke.

• Deep hands-on capability in Python and the modern AI stack: LLMs, RAG, agentic architectures, embeddings and vector retrieval, evaluation methodology, classical ML, and document AI/OCR.

• Strong cloud architecture depth across both Azure and AWS: Azure AI Foundry, Azure OpenAI, AI Search, Document Intelligence, AKS/Container Apps, PostgreSQL, Fabric — and AWS Bedrock, Knowledge Bases and Agents, Textract, SageMaker. Production experience on at least one is essential; credible working depth on both is what sets candidates apart.

• Track record operating in a regulated, audited environment where model decisions affect individuals and must be defensible.

• Proven ability to influence without authority: you have won over product, security, and compliance stakeholders and moved an AI agenda through enterprise approval gates.

• Strong vendor and partner management experience, including build-vs-buy decisions you have defended and lived with.

• Experience building and scaling technical teams in India, ideally within a GCC, global capability center, or captive setup.

Preferred

• Domain exposure to background screening, HR tech, healthcare, clinical or laboratory operations, or another high-volume document and compliance workflow business.

• Experience taking a GCC from delivery execution to product and platform ownership.

• Familiarity with AI governance and model risk management practice in an audited enterprise.
• Advanced degree in Computer Science, Machine Learning, Engineering, or a related field.

DISA Global Solutions is an equal opportunity employer committed to creating a diverse and inclusive workplace. All qualified applicants will receive consideration for employment without regard to race, color, religion, caste, gender, gender identity or expression, sexual orientation, national origin, age, disability, veteran status, marital status, or any other legally protected status. Any questions, please contact our DISA Recruiting Team (Careers-India <careers-india@disa.com>)