VP R&D

Position OverviewWe are seeking a VP of Research & Development to lead the transformation of our engineering organization into an AI-first, cloud-native delivery engine. This executive will drive the transition from a traditional software development lifecycle (SDLC) to a modern AI Development Lifecycle (AI-DLC) — embedding ML pipelines, intelligent automation, and data-driven decision-making into every stage of product development on Azure cloud infrastructure.This role combines strategic leadership with hands-on technical direction. You will own the R&D vision, architect scalable AI/ML systems, align technology investments with business outcomes, and build a high-performance team capable of executing at speed while maintaining trust, security, and reliability. The ideal candidate brings deep expertise in MLOps, cloud-native architecture, and GenAI integration — not just as concepts, but as production realities. Roles & ResponsibilitiesAI-DLC Transformation & Innovation • Lead the transition from SDLC to AI-DLC, integrating data pipelines, model development, evaluation, deployment, monitoring, and iteration into core engineering workflows • Design and implement end-to-end ML pipeline orchestration using Azure ML, MLflow, or Kubeflow — from feature engineering through model serving • Establish MLOps practices: model versioning, automated retraining triggers, A/B testing of models in production, drift detection, and performance monitoring • Define AI-DLC quality gating: automated checkpoints at each lifecycle stage (data validation, model evaluation, deployment readiness) • Implement continuous evaluation loops — production model performance monitoring against ground truth with automated rollback triggers • Build AI testing frameworks: unit testing for models, integration testing for pipelines, adversarial testing for robustness • Architect RAG pipeline designs including chunking strategies, embedding model selection, retrieval optimization, and reranking • Establish agentic AI design patterns — multi-agent orchestration, tool use, and autonomous workflow execution • Define prompt management systems: version-controlled prompt libraries, A/B testing of prompts in production • Lead fine-tuning vs. RAG vs. in-context learning decision frameworks — knowing when each approach fits • Implement LLM evaluation frameworks: hallucination detection, factual grounding, response quality scoring • Design guardrails and content filtering: input/output validation, toxicity detection, PII redaction in LLM responses • Drive synthetic data generation strategies for training and testing where production data is restricted • Architect vector database solutions and semantic search implementation (Pinecone, Azure AI Search, Weaviate) • Apply model compression and optimization for edge deployment — quantization, distillation, pruning • Optimize token cost and inference scaling strategies across AI workloadsTechnology Strategy & Cloud Architecture • In collaboration with Product, define and execute the long-term technology roadmap aligned to AI-native delivery • Architect Azure cloud-native infrastructure: AKS (Kubernetes), Azure Functions, Cosmos DB, Azure AI Services, Azure OpenAI Service • Drive Infrastructure as Code (Terraform, Bicep) and GitOps deployment models • Design API-first architecture and microservices patterns for scalable, real-time, AI-enabled platforms • Implement data pipeline architecture — ETL/ELT modernization from batch-heavy to event-driven, real-time enrichment using Azure Event Hubs / Kafka • Apply data mesh principles: domain-owned data products, federated governance, self-serve data infrastructure • Establish feature stores for ML — centralized, versioned, reusable feature engineering (Feast, Azure ML feature store) • Build data quality frameworks: automated schema validation, anomaly detection, lineage tracking • Design multi-region deployment strategies for global mobility use cases (latency, data residency, failover) • Lead edge computing and IoT integration strategies relevant to the company's device and hardware footprint • Implement FinOps discipline: cost modeling for AI workloads, GPU compute optimization, spot instance strategies, reserved capacity planning • Drive multi-modal AI capabilities — vision, document understanding, speech-to-text integrationProduct & Business Alignment • Partner with Product, Sales, and Customer Success to deliver impactful, customer-centric AI-powered solutions • Translate global mobility use cases into scalable technical solutions with measurable business impact • Align R&D priorities with company growth targets and market opportunities • Lead build vs. buy vs. partner evaluation frameworks specifically for AI capabilitiesGovernance, Security & Trust • Implement AI governance frameworks covering privacy, compliance, and model risk management • Apply Zero Trust security architecture to AI workloads — identity-based access, secrets management (Azure Key Vault), networ