AI Architect
The AI Architect is responsible for maintaining architectural coherence across GCS AI and Frictionless Time to Value (FTTV) initiatives, shaping use‑case and initiative shaping, backlog review, and initial solution definition before significant engineering effort is committed.
- Colombo, Western Province, Sri Lanka
- Posted Nov 12, 2026
INDUSTRYSoftware & InternetSaaS & Enterprise Software
Responsibilities
- Participate in the review and maintenance of the complete GCS AI backlog and the shaping of FTTV initiatives
- Assess whether proposed applications and FTTV solutions support GCS strategy, process direction, and the intended tooling landscape
- Identify duplicate, overlapping, unnecessarily isolated, or technically unclear proposals across the GCS AI backlog and FTTV initiatives before significant engineering effort is committed
- Work with business stakeholders and AI Engineering to refine use cases, clarify solution boundaries, and identify dependencies, integration points, and ownership
- Provide architecture input into GCS AI backlog and FTTV initiative decisions without assuming ownership of business prioritization
- Identify where GCS AI applications and FTTV solutions should reuse existing services, integrations, knowledge assets, evaluation approaches, or common components
- Work with AI Engineering to determine when a capability should be shared and when an application‑specific implementation is more maintainable
- Identify required connection points between solutions and ensure their purpose, ownership, and expected behavior are understood before implementation
- Review how structured knowledge is created, organized, accessed, and reused across AI applications and related FTTV solutions
- Promote consistent approaches to knowledge ownership, source traceability, permissions, versioning, update responsibility, and retirement
- Challenge solution‑specific knowledge stores when an appropriate governed source exists or is planned
- Work collaboratively with AI Engineering leads and developers to define practical architecture standards, reference patterns, and reusable approaches
- Distinguish between mandatory constraints, preferred patterns, implementation guidance, and examples
- Review how agreed standards are applied during solution shaping and propose changes when standards are unclear, outdated, or produce unintended results
- Identify and document justified exceptions, including their scope, consequences, dependencies, and support implications
- Present proposed exceptions for approval to Customer AI Experience leadership
- Determine which specialist perspectives are required for each proposed solution and involve them early enough to affect the design
- Ensure appropriate review of material decisions involving security, privacy, responsible AI, enterprise architecture, data, integrations, infrastructure, business processes, and product platforms
- Bring conflicting recommendations to the appropriate decision‑makers with the options and trade‑offs clearly stated
- Confirm that specialist recommendations are reflected in the proposed architecture or are addressed through a documented and approved exception
- Act as an architectural counterpart to AI Engineering leads and FTTV stakeholders during initiative shaping and initial solution definition
- Provide architectural direction without assuming responsibility for squad management or engineering delivery
- Work directly with developers when technical detail is required to evaluate a design, dependency, reuse opportunity, or proposed standard
- Use engineering and stakeholder feedback to test whether proposed standards and reusable approaches are practical
- Support consistency across initiatives and engineering squads without introducing an additional approval layer for routine implementation decisions
- Remain technically current enough to challenge assumptions and evaluate proposed designs credibly
- Prototype selected concepts, inspect implementations, or develop reference examples when needed to validate an architecture decision
- Maintain a current view of the GCS AI and FTTV solution landscape, including major boundaries, shared capabilities, dependencies, integrations, and knowledge relationships
- Document material architecture decisions, rationale, ownership, and approved exceptions at a level proportionate to the decision
- Assess whether proposed AI applications and FTTV solutions support the intended direction of GCS business processes and operational tooling
- Work with process, tooling, and functional owners to understand planned changes that may affect solution design
- Identify proposals that automate temporary workarounds, reproduce process defects, or create competing sources of process logic
- Make unresolved process or tooling dependencies visible before engineering commits to an approach
- Support a coherent GCS AI and FTTV solution portfolio that can evolve as GCS processes and tooling change without requiring each application or initiative to be independently redesigned
Requirements
Nice to have: enterprise AI applications in production, retrieval‑augmented generation, agent architectures, shared knowledge services, evaluation frameworks, AI observability, Azure AI services, Azure OpenAI, Azure AI Foundry, API‑first, event‑driven, microservice, cloud‑native, enterprise business applications, ERP
Skills
AI architecturelarge language modelsretrieval‑based solutionsagentic systemsstructured and unstructured knowledgemodern software architectureAPIsdistributed systemscloud servicesidentitysecurityintegration approachesarchitecture standardsreusable patternsmicroservicecloud‑nativeevent‑drivenAzure AI servicesAzure OpenAIAzure AI Foundryenterprise AI platformsAPI‑firstenterprise business applicationsERPAzure AI servicesAzure OpenAIAzure AI Foundrymicroserviceevent‑drivencloud‑native
Full description
Company details are for members
Sign up for a free job alert to see IFS's profile, website and LinkedIn, and the same for every employer on Referso Jobs.