About The Role
SportERP Limited is seeking a Senior Computer Vision Engineer to drive innovation and develop cutting-edge AI solutions for sports organizations worldwide. If you have 6–7 years of experience working with OpenCV, state-of-the-art computer vision models, and large language models, we want to hear from you. In this role, you'll leverage vector databases, document databases, and advanced machine learning frameworks to build robust image and video processing pipelines.
As part of our growing R&D; and Engineering teams in Pune, you'll collaborate closely with stakeholders in India and the UK, making excellent communication and English writing skills essential. This is your prospect to lead the development of next-generation computer vision tools in a fast-paced, SaaS-focused environment.
Responsibilities
Design, develop, and optimize computer vision pipelines using Python, OpenCV, and modern deep learning frameworks
Integrate large language models for advanced tasks such as image captioning, semantic understanding, and multimodal data processing
Utilize vector databases (e.g., FAISS, Milvus) and document databases (e.g., MongoDB, CouchDB) to store and retrieve embeddings efficiently
Collaborate with cross-functional teams (AI/ML engineers, product managers, etc.) to translate business needs into technical solutions
Implement scalable and secure data pipelines, ensuring high availability and low latency for real-time computer vision applications
Conduct research on emerging trends in AI and deliver innovative proof-of-concepts for sports technology solutions
Maintain detailed documentation and provide mentorship to junior team members
What We're Looking For
Required Skills & Experience
6–7 years of professional experience in computer vision and related fields
Expertise in Python and OpenCV for image/video processing
Hands-on experience with deep learning frameworks (e.g., TensorFlow, PyTorch)
Strong understanding of large language models and how th
📌 Senior Computer Vision Engineer (Pune)
🏢 SportERP
📍 Pune