Deputy Director - AI Solution and Platforms Observability
PepsiCo · Hyderabad
- Experience10–11 yrs
- SalaryNot disclosed
- Work modeonsite
- Levelexecutive
- Posted23 Sept 2026
About PepsiCo
PepsiCo is hiring in Hyderabad in consumer goods. This role looks for around 10+ years of experience.
Skills
- Computer Vision
- Vision-Language Models
- Vision Agents
- PyTorch
- OpenCV
- YOLO
- ONNX
- TensorRT
- NVIDIA DeepStream
- Kubernetes
- MQTT
- OPC UA
- Vision MLOps
The role
A Vision Intelligence architect works in an enterprise manufacturing company. They design, train, evaluate, and productionize Computer Vision models and multimodal Vision-Language Model solutions, build Vision Agents, and deploy real-time video analytics across edge GPUs, smart cameras, IoT systems, and cloud environments. Their skills include Computer Vision, Vision-Language Models, Vision Agents, PyTorch, OpenCV, YOLO, ONNX, TensorRT, NVIDIA DeepStream, Kubernetes, MQTT, OPC UA, and Vision MLOps.
Full job description
Overview
Overview:
Senior hands-on individual contributor responsible for enabling enterprise Vision Intelligence capabilities across Computer Vision, Vision-Language Models (VLMs), Vision Agents, Edge AI and IoT. Partners with Domain and Enterprise Architecture teams to translate business needs and architecture patterns into scalable, production-ready capabilities.
Responsibilities
Key Responsibilities:
Serve as the technical SME for Vision Intelligence, covering CV, VLMs, Vision Agents, Edge AI and IoT.Design, train, evaluate and productionize CV models for detection, classification, segmentation, tracking, OCR, anomaly detection and activity recognition.Develop VLM and multimodal solutions for image/video understanding, visual reasoning and operational intelligence.Build Vision Agents combining visual intelligence, IoT telemetry, enterprise context, tools and workflows.Enable reusable platform capabilities including inference services, APIs, SDKs, model pipelines and reference implementations.Deploy and optimize AI workloads across edge GPUs, industrial PCs, smart cameras and cloud/edge environments.Integrate cameras, sensors and IoT/OT systems using RTSP, ONVIF, MQTT, OPC UA and related technologies.Establish Vision MLOps practices for datasets, evaluation, CI/CD, deployment, monitoring and model drift.Partner with Manufacturing, Warehouse, Quality, EHS and other domains to move solutions from POC → Pilot → Production → Scale.Evaluate emerging CV, VLM, Edge AI, Vision Agent and Physical AI technologies through hands-on prototyping and benchmarking.Collaborate with Domain and Enterprise Architects and mentor engineering teams on Vision Intelligence implementation.
Qualifications
Qualifications:
Overall 10+ years of overall experience is required.Bachelor’s degree in Computer Science, Engineering, or related field (or equivalent experience)8–12+ years of AI/software engineering experience, including 5+ years in Computer Vision/Vision AI.Strong hands-on experience developing and deploying production-grade CV models and real-time video analytics.Practical experience with VLMs, multimodal AI and Vision Agents.Experience with PyTorch, OpenCV, YOLO, ONNX, TensorRT, NVIDIA DeepStream or equivalent technologies.Experience deploying AI on GPU/edge platforms, containers and Kubernetes/K3s.Strong IoT/OT integration experience with cameras, sensors, gateways and industrial environments.Experience with CV/Vision MLOps, model optimization, evaluation and production monitoring.Demonstrated experience taking Vision AI solutions from prototype through production at multiple sites.Manufacturing, warehouse, supply-chain, quality, EHS, robotics or industrial automation experience is highly desirable.Exposure to Digital Twins, sensor fusion, robotics and Physical AI is a plus.