Sai Nivedh Vallala Email & Phone Number

Bengaluru, Ka, In

Hyderabad, In

Chennai, Tamil Nadu, India

• Work closely with the Founder, Lead to Implement the Product & Technology Roadmap.• Guide and work along with the AI / ML team on activities & R&D performed in platform.• Developing Vertical Solutions with platform to support team in Presales, showcase to customers• R&D on Planogram Compliance check and insights on Share of Shelf of SKUs displayed in retail stores. • Developed Stiching and Cluster engines to automate and aid Planogram Compliance check on Image/Video through mobile video, Shelf Camera, Dome Cameras.• Trained Unsupervised pipeline over 2 million SKU annotated Datasets with 110K classes.• Semi Supervised approach for Classifying any number of SKUs across Categories.• Leveraged Feature Store for low latency classification of items/categories/sku with over 99% F1 score.

Chennai, Tamil Nadu, India

• Brainstorming & Requirement gathering for the Product Engines and AI Implementations.• Work closely with backend Design team on API /Integration design for Coherent development inparallel• Implemented Active Learning pipeline to reduce cost and time of annotators, making the labelling process 15% more efficient and faster.• Developed Model Builder engines, distributed training workflows for Object Detection, Semantic/Instance Segmentation problems covering variety of industry usecases suchas • Retail - Dense OD, Occlusion • Defense - Satellite Imagery, Tiny, Rotated OD • Visual Inspection - High Accurate, Fast OD• Added Vector Search Engine to Platform by Analyzing different cutting-edge Vector databases and developed a search pipeline framework which is 10X faster than Elasticsearch• Developed a generative pipeline using Mixture Models, VAEs to create new content by learning underlying Image distribution, enabling users to add or augment data to improve model performance. • Create a Search engine over Knowledge Base of Insurance companies for quick access to Customer Success team in Insurance industry.• The pipeline includes finetuning over LLMs for Embedding retrieval and the AUC-ROC was 98.6%

Banglore

Master Of Technology - Mtech, Aerospace Engineering

Bachelor Of Technology - Btech, Aerospace Engineering