Software Engineering Specialist - Agentic AI

Job description

Req ID: 61820 Job Function: Software Engineering Posting Start Date: 13/08/2026 Posting End Date: 19/08/2026 Division: Digital Job Location: IND-Bengaluru-RMZ Ecoworld Advertised Salary: Competitive Job Req ID: Posting Date: Function: Location: Salary: ## **About the role** The Software Engineering Specialist independently executes advanced activities to deliver the engineering strategy and roadmap that supports BT's commercial strategy through cross functional business partnering and the participation of a team that pursues innovation as well as engineering excellence. ## **What you’ll be doing** 1) Engineering Strategy, Predictability & Measurability - Deliver the AI Factory engineering strategy so teams can build high-quality agentic solutions with predictable outcomes (capacity planning, delivery metrics, quality gates, cost-to-serve tracking, platform SLAs). 2) Technical Decisions & Architecture Across Platforms - Make complex technical decisions spanning Azure OpenAI model choices, orchestration patterns, Copilot Studio design, ServiceNow integration patterns, and enterprise architecture. 3) Solve Strategic/Complex Problems with Leading-edge Solutions - Resolve complex agentic issues: prompt injection, tool misuse, grounding failures, hallucinations, latency spikes, knowledge freshness, agent memory pitfalls, and secure tool execution at scale. 4) Execute & Contribute to the Technical Roadmap - Define and execute a roadmap for: o Azure OpenAI capability adoption (models, embeddings, content filtering, caching) o Copilot Studio extensibility (connectors, actions, plugins) o ServiceNow AI experiences (Virtual Agent/Now Assist patterns, workflow triggers) o Shared runtime components (tool registry, policy engine, evaluation services) 5) Engineering & Operational Excellence (Metrics + Improvement Loops) - Establish excellence practices: definition-of-done for AI, release criteria, evaluation regression suites, security reviews, performance baselines, and runbooks. 6) Foster Innovation with High Reliability - Create a culture of rapid experimentation with controls: safe sandboxes, feature flags, A/B tests, prompt/version governance, and production readiness checklists. 7) Hands-on Coding, Testing & Reviews - Write, test, and review code across agent services, middleware, connectors, and orchestration logic; refactor prompt flows and agent policies as required. 8) Resolve Escalations (Deep Technical Troubleshooting) - Debug and troubleshoot across: o Agent orchestration services o Prompt workflows + evaluation failures o ServiceNow integrations / APIs o Copilot Studio action chains o Identity/access issues (Entra ID) o Observability traces and incidents 9) Drive Technical Vision & Innovation - Contribute to the broader technical direction: new patterns for agent planning/routing, safe tool calling, RAG design, memory strategies, and cross-platform integration. 10) Tooling & Automation for Developer Productivity - Implement and maintain CI/CD and automation: o Prompt + flow versioning o Automated eval pipelines o Quality gates (groundedness, relevance, toxicity checks) o Automated release validation o Developer templates and scaffolding 11) Architectures & Standards for Enterprise Scale - Define enterprise standards for: o Agent runtime design o RAG architecture (indexing, retrieval, citations) o Secure tool execution patterns o Data access boundaries o Tenant-level governance + audit logging o Multi-environment promotion (dev/test/prod) 12) Build New Software + Data-driven Improvements (Reduce Tech Debt) - Research, design, and build new components, and perform deep analysis of agent telemetry to reduce tech debt and improve reliability, performance, and developer experience. 13) Mentorship & Technical Coaching - Mentor engineers and squads via design reviews, code reviews, pairing sessions, office hours, and playbooks. 14) Knowledge Leadership & Emerging Trends - Continuously research and share best practices in agentic AI, LLMOps, Responsible AI, evaluation techniques, and platform feature evolution. ## **Essential Skills / Experience** Technical (Must-have) - Strong experience building LLM-powered solutions (RAG, tool calling, prompt engineering, evaluation-driven development). - Strong engineering background in Python and/or TypeScript/C# with production-grade quality. - Experience with Azure components relevant to AI workloads (APIs, security, monitoring, pipelines). - Ability to design scalable enterprise architectures and integration patterns. Platform (Must-have for this role) - Hands-on experience with Azure OpenAI (model usage patterns, embeddings, filtering/safety, cost optimisation). - Experience building copilots/assistants using Copilot Studio (connectors/actions/flows). - Experience integrating with ServiceNow (workflows, APIs, automation patterns; AI/Virtual Agent exposure is a plus). Engineering Excellence - CI/CD, automated testing, code review discipline, documentation playbooks. - Observability and production support mindset (SLOs, incident response, postmortems). Security & Governance Mindset - Understanding of identity/access controls, secure integration, and AI risk controls. - Ability to implement guardrails that prevent unsafe or non-compliant agent behaviour. ## **Desirable Skills / Experience** - LLMOps/AgentOps tooling experience (evaluation pipelines, prompt versioning, tracing frameworks). - Knowledge of Microsoft governance/security ecosystem (e.g., Entra ID, Purview concepts, SOC/monitoring practices). - Experience implementing knowledge ingestion pipelines and vector stores. - Experience with ServiceNow CoE operating models and enterprise platform governance. BT Group is the UK’s leading communications group and the holding company behind some of the country’s most recognised brands – in