Nivedha Thangaraju

Recent AI Master's graduate from the University of Surrey, skilled in computer vision and deep learning, with a solid foundation of over 3 years in software development. Seeking an AI/ML engineering role to contribute innovative solutions, leverage skills refined through diverse projects, and drive continuous growth and impact within a dynamic environment.

AI Engineer (June 2023 - Current)
Command Rooms, Norwich, UK

• Led the development of specialized third-party AI plugins tailored for camera manufacturers.
• Collaborated to seamlessly integrate these plugins with Network Optix, enhancing object detection and motion detection capabilities.

Software Engineer (September 2018 - January 2022)
Cisco Systems, Bangalore, India

• Played a key role in designing and developing contact center software solutions, with a focus on ASR/TTS systems and NLP integration into conversational IVR systems within the Collaboration Technology Group.
• Created algorithms to optimize Voice Activity Detection, resulting in resource savings by preventing unused channel usage.
• Orchestrated the development of essential microservices for a smooth transition to cloud infrastructure, dockerizing critical solution components.
• Engineered a DTMF input encryption feature for PCI compliance, safeguarding customers' personally identifiable information (PII).
• Improved system serviceability by implementing advanced features and statistical analytics for Call Volume analysis.
• Established a robust security framework for seamless integration with cloud services, enhancing data protection and system integrity.

Master of Science in Artificial Intelligence (2022 - 2023)
University of Surrey

• Grade: Distinction
• Relevant Modules: Computer Vision and Pattern Recognition, Image Processing and Deep Learning, AI and AI Programming, Speech and Audio Processing and Recognition, Ethics and Regulation of AI.
• Thesis: "Audio Retrieval by Natural Language Queries" supervised by Dr. Wenwu Wang.
  • Designed an audio retrieval system using transformer encoders based on ViT and BERT to represent audio and text modalities, respectively.
  • Optimized the cross-modal joint embedding space with Softmax-based loss functions, achieving benchmark-beating performance on several metrics for AudioCaps and Clotho datasets.

Bachelor of Engineering in Electronics and Communication (2014 - 2018)
PSG College of Technology, Coimbatore