Omkar Kumbhar Email & Phone Number

Hyderabad, Tg, In

Redmond, Washington, Us

pipelines, ML, GenAI and core web search engines

New York, Ny, Us

Startup generalist: Hands-on Engineer, data guy, growth and data strategy, leadership, and whatever it takes to keep Skylyte growing, and successful with maximum impact on reducing burnout and increasing resilience in people. Domain:Speech Processing, NLP, LLM, Computer Vision- Leading speech data collection for workplace wellbeing and burnout - aurora.skylyte.com- Researching, prototyping, and building ML modules for speech processing for Skylyte on AWS Cloud- Using cutting edge developments in speech processing for doing internal research on effects of Burnout risk on vocal biomarkers over time- Taking care of data strategy and growth for Skylyte- Scaling ML APIs on sentiment analysis, topic modeling, and affect recognition from text and audio- Personalization, simplication and recommendation module for combining dashboard data into bite sized pieces of next-steps using Prompt Engineering and GPT3.5 turbo

New York, Ny, Us

Skylyte is the modernizer of employee engagement, and helping teams teams navigate burnout risk and mental health.- Research in comparing self reported moods between users with predicted mood from voice- Developing toolkits capable of providing end user an overview in their work life based on voice reflections and burnout

New York, Ny, Us

Supervisors: Prof. Denis Pelli, Prof. Elena Sizikova- Assisted at Pelli Labs by providing support and insights in vision and object recognition.- Assisted in research related to comparing human and machine vision on noisy reading task.- Lead a project in comparing humans and machine vision with respect to reaction time by using Pytorch and architectures like CRNN, Seq2Seq and MSDNet.- Created a benchmark for Speed Accuracy Trade-Offs in Humans and compared models like MSDNET, SCAN, and rCNN models to model human reaction time. https://human-rt.netlify.app/- Worked as a MATLAB Programmer to design experiments related to neuroscience and vision and contributing to Psychtoolbox - http://psychtoolbox.org/

Frankfurt, Hesse, De

Innoplexus is an AI and Blockchain-powered company providing real-time data analytics to clients in Pharma, Life Sciences and Financial Services.Domains: Computer Vision | Natural Language Processing - Working in the Innovation department, applying theoretical concepts to build products around deep learning and ML models. - Primarily used Pytorch, Keras, Tensorflow and NVIDIA GPUs to design, scale and run applications in computer vision for Innoplexus.- Worked on patented AI-based software for Innoplexus - PDF Extraction using computer vision.- Created a PDF Table Extraction module using computer vision techniques for data extraction.- Implemented traditional NLP based module for finding transitions in text to find and predict patterns in research papers. - Understanding requirements from SMEs from life-sciences and bio-technology to make scalable AI-based utilities.

Castalia Research Labs was a company providing services in image processing, machine vision, optics and machine learning for clients in the automotive and textile industry.Domain:Computer Vision- Training in deep learning for computer vision by referring to Stanford University's CS231n - http://cs231n.github.io/convolutional-networks/- Worked in developing defect detection system in elastic bands using custom CNN models.

Master Of Science - Ms, Computer Science

Bachelor Of Engineering - Be, Electronics And Telecommunications