Expert
We are looking for a Product Architect for our Image recognition solution, responsible for the end-to-end technical architecture of the product. You will bridge product strategy and engineering execution, defining how the product scales across markets, integrates with enterprise systems, and evolves with AI/CV innovation. This role works in close partnership with the Global Lead, mobile/SDK engineering team, Backend engineering, data science teams, and S&T leadership. You will own architecture decisions across mobile, backend, cloud infrastructure, ML inference, and data pipelines — ensuring the platform meets global security, compliance, and performance standards while remaining agile enough to support continuous market deployments.
Define and maintain the end-to-end technical architecture for the IR product across mobile, API, ML pipeline, and cloud layers
Design scalable, multi-region cloud infrastructure on Azure (API Management, Blob Storage, Azure Functions, Cosmos DB, KubeRay/Ray Serve)
Lead architecture decisions for mobile app features, edge/cloud validation modes, and offline-first mobile scenarios
Ensure architecture supports high-volume concurrent field usage for global markets
Define data contracts, API schemas, and integration patterns between mobile app, SDK, backend services, and Salesforce apps
Architect the ML inference pipeline for computer vision models including SKU detection, planogram compliance, and share-of-shelf measurement
Design model deployment and versioning strategies using KubeRay/Ray Serve or equivalent orchestration platforms
Collaborate with data science teams to establish model accuracy baselines, A/B testing frameworks, and continuous learning pipelines
Evaluate and integrate emerging AI capabilities — including LLMs, VLMs, and agentic AI — into the IR Product platform roadmap
Define iOS and Android architecture standards for geo-fencing, image capture, local caching, and sync behaviour
Establish mobile contractor technical standards, code review frameworks, and platform-specific constraints documentation
Design Salesforce integration patterns for field rep workflows, task assignments, and audit result reporting
Architect authentication, session management, and device-level security controls for global field deployments
Own infrastructure architecture for reliability, fault tolerance, and disaster recovery — including GCS fault tolerance and Redis integration for KubeRay head pod stability
Define storage architecture and lifecycle policies to address accumulation issues (zip files, image blobs, log data) at scale
Conduct root cause analysis (RCA) for critical infrastructure incidents and produce executive-grade escalation documentation
Ensure infrastructure architecture meets GDPR, India DPDP Act, and market-specific data privacy regulations for GPS and biometric data
Design privacy-by-design data flows for GPS location data, field rep identifiers, and store imagery in compliance with GDPR, DPDP, and other applicable frameworks
Define data residency, data minimisation, and retention policies across global cloud deployments
Establish security architecture standards for mobile app communications, API authentication, and cloud storage access controls
Produce clear, executive-facing architecture documentation, TRDs (Technical Requirements Documents) in partnership with the Global PM
Present architecture proposals and trade-off analyses to leadership and cross-functional stakeholders
Mentor Engineering team on product architecture standards and best practices
Maintain a living architecture registry — covering system diagrams, ADRs (Architecture Decision Records), and integration maps
12-15+ years in software/product architecture roles, with at least 3 years in mobile-first or AI/ML-integrated platforms
Proven experience architecting cloud-native platforms on Azure (API Management, Functions, Blob Storage, Cosmos DB)
Hands-on experience with ML inference infrastructure — KubeRay, Ray Serve, TorchServe, or equivalent
Strong background in mobile architecture for iOS and/or Android, including offline-first design patterns
Experience with computer vision, image recognition, or retail execution platforms strongly preferred
Prior exposure to FMCG, CPG, or retail technology platforms is a significant advantage
Cloud: Azure (primary), AWS/GCP (secondary) — infrastructure design, cost optimisation
Mobile: iOS (Swift/Objective-C), Android (Kotlin/Java), cross-platform considerations (React Native, Flutter)
ML/AI: Model serving pipelines, MLOps practices, CV model deployment (object detection, classification)
Integration: REST APIs, event-driven architectures (Kafka, Service Bus)
Data: Cosmos DB, PostgreSQL, Data bricks, cloud storage lifecycle management, data governance frameworks
DevOps: Kubernetes, Docker, CI/CD pipelines, infrastructure monitoring and alerting
Exceptional ability to translate complex technical architecture into clear, business-relevant narratives for executive audiences
Strong cross-functional collaborator — equally effective with product managers, data scientists, field operations leads, and compliance teams
Demonstrated ownership mindset — proactively identifies risks, escalates with solutions, and drives resolution end-to-end
Comfortable operating in ambiguous, multi-market environments with competing priorities
Excellent written communication — capable of producing publication-quality TRDs, PRDs, and architecture decision records
Sign up to apply and find out right away if you're a fit.
Your agent will tell you — in seconds.
Sign up and I'll tell you right away how well PepsiCo matches you — what you already have, and what's missing. Then I stay on it: I search for you and only write when I find something worth your time.