Computer Vision and Edge AI Intern

Develop and harden computer vision and edge AI systems for crop protection, optimizing models for real-world conditions and constrained hardware.

🎯 About the Role

We're looking for someone to build and harden the computer-vision and edge-AI that powers Kavach - taking detection and on-device decision-making from "works" to "works reliably, in the real world, on constrained hardware." This is a deep, hands-on engineering role: you'll own real problems end-to-end, from data to model to the device itself.

🛠️ Key Responsibilities

• 🏗️ Build and improve datasets and data pipelines for multi-species wildlife detection.
• ☁️ Train and tune object-detection models (YOLO family) on cloud training platforms.
• 💻 Optimise and deploy models to edge AI hardware.
• 🛡️ Harden detection and on-device decision logic for real-world conditions.
• 🧠 Develop on-device logic that decides how the system responds.
• 🛠️ Build evaluation and benchmarking tooling.

✅ Requirements & Skills

• 🐍 Solid Python - clean, maintainable code.
• 💡 Working knowledge of computer-vision / ML fundamentals; familiarity with YOLO / PyTorch / Ultralytics is a strong plus.
• ☁️ Hands-on experience with at least one cloud/ML training platform (AWS SageMaker or Google Colab).
• 🐧 Comfort on Linux and the command line.
• 🗄️ Git in your daily workflow.
• 📊 Careful, methodical data handling.
• 🛠️ Demonstrated ability to unblock yourself using documentation and AI tools.
• 📍 Available in person at IISER Mohali for the full 3–6 months.

🎁 Perks & Benefits

• 💰 Competitive Stipend
• 💻 Serious tooling and resources (NVIDIA, AWS, Zoho).
• 🛡️ Government-backed deep-tech.
• 🤝 Real ownership and direct mentorship from the founder.
• 🚀 Path to full-time for interns who deliver.

⏳ Duration: 3-6 Months

📍 Location: IISER Mohali