Project Lead – Edge AI and Computer Vision

TATA ELXSI; Bengaluru; On-site; INR 2,500,000 - 4,500,000; Full time

Job summary

Tata Elxsi, based in Bengaluru, India, seeks an expert in Edge AI and computer vision to lead end-to-end delivery of advanced AI solutions on embedded and on-premise platforms. The role focuses on hardware-aware optimization and real-time performance within constrained environments.

Candidates will drive GenAI and agent-based workflows, coordinate with IoT integrations, and work closely with system architects to meet AI KPIs in an agile/hybrid setting.

Qualifications

• Deep expertise in CNNs, object detection, semantic and instance segmentation, and vision transformers.
• Experience deploying models on resource-constrained edge devices (<4GB RAM, <10W) and optimizing for edge inference.
• Proficiency with edge AI frameworks and hardware platforms (Jetson, Movidius, Raspberry Pi, NPUs).
• Knowledge of Generative AI (GANs, diffusion, VAEs) and reinforcement/agentic AI concepts for edge environments.

Responsibilities

• Lead end-to-end delivery of Edge AI, CV, and Generative AI solutions.
• Drive architecture and delivery planning for AI on edge devices and on-premise infrastructure.
• Collaborate with architects to design hybrid AI systems with local and centralized components.
• Oversee optimization for model size, latency, throughput, and accuracy.
• Coordinate integration with cameras, sensors, and IoT devices.
• Drive Agile / Hybrid delivery models and ensure safety/regulatory compliance where applicable.

Skills

CNNs

Object detection

Semantic segmentation

Instance segmentation

Vision transformers

Edge Deployment

Model Optimization

Quantization

Pruning

Knowledge distillation

Neural architecture search

Edge Frameworks

Hardware Platforms

Image Processing

Generative AI

Agentic AI

Reinforcement learning

Education

Bachelor's degree in Computer Science or related field

Master's degree preferred

Tools

TensorFlow Lite

ONNX Runtime

OpenVINO

TensorRT

PyTorch Mobile

TVM

Job description

Opportunities unlimited for everyone - be bold, curious, and seek to shape the future. Explore what's possible, discover what you love to do, and find accelerated paths for growth. At Tata Elxsi, You Matter!

Tata Elxsi is among the worlds leading providers of design and technology services across industries, including automotive, broadcast, communications, healthcare, and transportation. Tata Elxsi works with leading OEMs and suppliers in the automotive and transportation industries for R&D, design, and product engineering services from architecture to launch and beyond.

Key Responsibilities:

• Lead end-to-end technical delivery of Edge AI, Computer Vision, and Generative / Agentic AI solutions
• Drive architecture and delivery planning for AI systems deployed on edge devices and on-premise infrastructure
• Guide implementation of:
  
  • Computer Vision and image processing pipelines
  • Deep Learning models optimized for edge inference
  • Generative AI and agent-based workflows for local reasoning, decision-making, and automation
  • Sensor integration and real-time data acquisition
• Collaborate with architects to design hybrid AI systems where GenAI/agents operate locally or in coordination with centralized services
• Ensure optimization across:
  
  • Model size, inference latency, throughput, and accuracy trade-offs
  • Efficient execution of GenAI models and agent logic on constrained platforms
  • Platform-specific acceleration (CPU, GPU, NPU, DSP)
• Drive techniques such as quantization, pruning, distillation, and hardware-aware optimization
• Oversee benchmarking, performance tuning, and real-time validation
• Lead deployment of AI solutions on embedded, edge, and on-premise platforms
• Coordinate integration with cameras, sensors, industrial devices, and IoT systems
• Translate domain needs into technical delivery plans, AI KPIs, and system constraints
• Communicate risks, trade-offs, and performance expectations to stakeholders
• Own delivery plans, schedules, dependencies, and technical risks for complex AI programs
• Drive Agile / Hybrid delivery models supporting hardware-software co-development
• Ensure compliance with safety, regulatory, and quality standards where applicable

Required Technical Skills & Experience:

• Computer Vision: Deep expertise in CNNs, object detection (YOLO, SSD, Faster R-CNN), semantic segmentation (U-Net, DeepLab), instance segmentation (Mask R-CNN), and vision transformers
• Edge Deployment: Extensive experience deploying models on resource-constrained devices with <4GB RAM, <10W power budgets
• Model Optimization: Hands-on experience with quantization (INT8, FP16), pruning, knowledge distillation, and neural architecture search
• Edge Frameworks: Proficiency with TensorFlow Lite, ONNX Runtime, OpenVINO, TensorRT, PyTorch Mobile, TVM, or similar
• Hardware Platforms: Experience with NVIDIA Jetson (Nano, TX2, Xavier, Orin), Intel Movidius/NCS, Raspberry Pi, ARM Mali, Qualcomm NPUs
• Image Processing: Strong foundation in OpenCV, PIL/Pillow, scikit-image, traditional CV algorithms, and image enhancement techniques
• Generative AI: Knowledge of GANs, diffusion models, and VAEs for synthetic data generation and edge-based generation
• Agentic AI: Understanding of reinforcement learning, decision-making systems, and autonomous agents for edge environments

Domain Knowledge:

• Manufacturing: Understanding of industrial automation, machine vision systems, quality control processes, and factory standards (ISO 9001)
• Medical Diagnostics: Familiarity with medical imaging modalities (X-ray, CT, MRI, ultrasound), DICOM standards, FDA regulatory pathways, and clinical workflows
• Automotive: Knowledge of ADAS systems, autonomous driving stacks, automotive sensors (cameras, LiDAR, radar), and functional safety (ISO 26262)
• Experience in at least one vertical with deep understanding of industry requirements and use cases

Technical Infrastructure:

• Embedded Systems: Understanding of embedded Linux, RTOS, device drivers, and low-level optimization
• Sensor Integration: Experience with camera interfaces (CSI, USB, MIPI), sensor protocols (I2C, SPI, CAN), and multi-sensor fusion
• Edge Computing: Knowledge of edge-cloud architectures, fog computing, and distributed inference strategies
• MLOps for Edge: Experience with edge-specific CI/CD, containerization (Docker on edge), and OTA update mechanisms
• Programming: Proficiency in Python, C/C++ for performance optimization, and CUDA/OpenCL for GPU acceleration

Preferred Qualifications:

• Bachelor's degree in Computer Science, Software Engineering, or related technical field; Master's degree preferred
• Exposure to embedded AI accelerators and edge platforms
• Familiarity with safety-critical or regulated environments