Experience

Couture.ai

• India
• Software Development
• 1 - 100 Employee

• Lead Deep Learning Engineer

  • Apr 2022 - Present

  Bangalore Urban, Karnataka, India Current: Building business LLMs. Open to discuss. Let's connect. Video Dubbing in Multiple Indic Languages • Orchestrated the development of a comprehensive video dubbing pipeline enabling seamless dubbing from English to diverse Indic languages. • Elevated the quality of synthesized speech, surpassing typical robotic-sounding output from Speech Synthesis models. • Leveraged Speech Style Transfer techniques to deliver impeccable dubbing in the desired… Show more Current: Building business LLMs. Open to discuss. Let's connect. Video Dubbing in Multiple Indic Languages • Orchestrated the development of a comprehensive video dubbing pipeline enabling seamless dubbing from English to diverse Indic languages. • Elevated the quality of synthesized speech, surpassing typical robotic-sounding output from Speech Synthesis models. • Leveraged Speech Style Transfer techniques to deliver impeccable dubbing in the desired source voice. • Preserved the integrity of original background music during the dubbing process. • Enhanced dubbing capabilities by enabling replacement of text-headlines in the target language, specifically tailored for news videos. • Augmented translation quality through the utilization of a self-hosted LLM model, resulting in natural and context-aware dubbing. • Continuously exploring avenues to refine prosody in speech synthesis for an elevated user experience. • Integrated our in-house Platform Video Studio, empowering the ability to fine-tune dubbing parameters, encompassing Transcripts Generation, Speech Generation, and OCR In-painting Generation. • For further information, please feel free to reach out to us at: https://caiplatforms.com/contact.html Video Trailer, Summary, GIF, Thumbnail Generation • Led the development of an advanced trailer generation module, leveraging diverse video summarization methods. • Empowered users with meticulous control over video generation through our cutting-edge Platform Video Studio. • Customized trailer generation options, enabling the selection of different summarization methods, adjusting duration, frame rate, and default thumbnail selection. • Seamlessly created captivating GIFs with configurable parameters, including frame rate, duration, and thumbnail customization. • For additional insights, please contact us at: https://caiplatforms.com/contact.html

• Deep Learning Engineer III

  • Apr 2021 - Mar 2022

  Bangalore Urban, Karnataka, India Content Moderation for News Channel Videos • Led the comprehensive development of the end-to-end pipeline, encompassing seamless ingestion and moderation for both Livestream and Offline Videos. • Developed robust models to accurately tag violent and explicit scenes in News videos across various Indic languages. • Implemented violence detection, including crime scenes (such as gun-shooting, fighting, stone-pelting, road accidents, arson, etc.). •… Show more Content Moderation for News Channel Videos • Led the comprehensive development of the end-to-end pipeline, encompassing seamless ingestion and moderation for both Livestream and Offline Videos. • Developed robust models to accurately tag violent and explicit scenes in News videos across various Indic languages. • Implemented violence detection, including crime scenes (such as gun-shooting, fighting, stone-pelting, road accidents, arson, etc.). • Ensured compliance by detecting and flagging the presence of minors in livestream videos. • Employed explicit content detection to identify and tag nudity, bloody scenes, and gory visuals. • Significantly reduced News story approval time through automated moderation, minimizing reliance on manual interventions. • Leveraged our in-house Platform Training Feedback Loop to continually enhance accuracy and mitigate false positives. • For further details, please don't hesitate to contact us at: https://caiplatforms.com/contact.html

• Deep Learning Engineer I

  • Apr 2020 - Mar 2021

  Bangalore Urban, Karnataka, India Video Attributes Enrichment • Developed a hybrid recommendation pipeline for a video serving platform (similar to youtube). • Built the video embeddings engine : key scenes, key frames, audio and video level features. • Achieved 9x runtime improvement of the pipeline by using Video Summarization and Parallelization. Portfolios Showcase • Developed several proof of concepts : Multilingual Speech to Text*, Fashion Clothings Parsing*, Video Background Blur and Swap etc. which led… Show more Video Attributes Enrichment • Developed a hybrid recommendation pipeline for a video serving platform (similar to youtube). • Built the video embeddings engine : key scenes, key frames, audio and video level features. • Achieved 9x runtime improvement of the pipeline by using Video Summarization and Parallelization. Portfolios Showcase • Developed several proof of concepts : Multilingual Speech to Text*, Fashion Clothings Parsing*, Video Background Blur and Swap etc. which led to multiple clients acquisition*.

• Associate Data Scientist

  • Jan 2019 - Mar 2020

  Bangalore Urban, Karnataka, India • Worked on Brands Detection (~300 popular brands) and Visual Similarity for a large E-commerce client. • Worked on Color Extraction from clothing Images. It helps in finding similar color items. • Integrated Tensorflow and Pytorch Serving into the Organisation’s Data Science Platform. • Setup gpu based training and automation processes for cloud and on-premise servers.



Ittiam Systems Pvt Ltd

• India
• Software Development
• 100 - 200 Employee

• Data Science Engineer

  • Jul 2017 - Dec 2018

  Bangalore Urban, Karnataka, India I worked as Engineer in the Corporate RnD Team for IttiamInsite-AI-VisualAnalytics-Retail-Insights product. • Worked on porting trained neural network models from Tensorflow to Android Bazel Build and profiled multiple CNNs for performance on an ARM processor. • Built a demo interfacing a USB web camera that performs face detection using dlib library and then calls an Ittiam trained CNN for age, gender, and ethnicity estimation. • Worked on building a similarity based retrieval of… Show more I worked as Engineer in the Corporate RnD Team for IttiamInsite-AI-VisualAnalytics-Retail-Insights product. • Worked on porting trained neural network models from Tensorflow to Android Bazel Build and profiled multiple CNNs for performance on an ARM processor. • Built a demo interfacing a USB web camera that performs face detection using dlib library and then calls an Ittiam trained CNN for age, gender, and ethnicity estimation. • Worked on building a similarity based retrieval of similar clothing items from a catalog based on an image containing a clothing item and hosted it on Google cloud as a demo. • Worked on building a CNN for classifying Japanese vs. non-Japanese faces. • Worked on building a CNN for classifying emotions in face images. • Automated web page scraping for neural network training. • Built a semi-automatic annotation tool for a top-down stereo camera based person re-identification use-case. Show less



Wipro

• India
• IT Services and IT Consulting
• 700 & Above Employee

• Software Engineering Intern

  • May 2016 - Jun 2016

  Bangalore Urban, Karnataka, India • Worked on a DevOps project.



Cretif Safety Solutions

• Winter Intern

  • Dec 2015 - Jan 2016

  IIT Gandhinagar • Android app development under which I mainly worked on socket io, http connection , and used to exchange data (streaming as well) from/to the AWS server. • Apart from it, I also worked on frontend technology like materialize, customisation of wordpress and created custom themes,plugins and widgets for the company blog.



National Institute of Technology Surat

• Higher Education
• 1 - 100 Employee

• Summer Research Intern

  • May 2015 - Jun 2015

  Surat • Research Topic : Extended vertical FP growth algorithm on columnar database.