Aniket Mohanty

Aniket Mohanty

About

Detail

Founder
New York, United States

Timeline

work
Job
school
Education

Résumé

Jobs verified_user 0% verified
  • Coral
    Founder
    Coral
    Jul 2024 - Current (1 year 11 months)
  • LimeChat
    ML Lead
    LimeChat
    Oct 2022 - Jul 2024 (1 year 10 months)
    Introduce ChatGPT in ECommerce chatbot: - Spearheaded the integration of ChatGPT to fill gaps in existing bot responses and tailored content for a WhatsApp audience, resulting in a 20% uplift in automation and user engagement. Create a framework to train transformer models - Established a Django server with Celery for queuing training jobs and led the development of a UI for model training, allowing anyone to schedule training via the UI and reducing training time by 80%.
  • LimeChat
    SDE-2 (NLP)
    LimeChat
    Jun 2021 - Oct 2022 (1 year 5 months)
    Chatbot Performance Monitoring & Smart Bug Reporting System: - Designed the HLD and LLD, streamed data to Elasticsearch, and set up Kibana dashboards & alerts, enabling proactive bug resolution and identification of chatbot response time bottlenecks. Improve browsing experience on chat for apparel brands: - Automated product attribute mining from descriptions and introduced ML algorithms for product search, enhancing UX for product exploration. This increased product discoverability by 40% and boosted sales by 10%.
  • Aina
    Computer Vision Research Engineer
    Aina
    Oct 2020 - Apr 2021 (7 months)
    Key Achievements: - Developed a virtual clothing try-on application from inception, enabling real-time body measurements and 2D garment visualization. - Achieved an impressive measurement accuracy rate of 96.2%. - Successfully launched the application on TestFlight for iOS.
Education verified_user 0% verified
  • IIIT Hyderabad
    Bachelor of Technology - BTech, Electronics and Communications Engineering, Electronics and Communications Engineering
    IIIT Hyderabad
    Jan 2017 - Jan 2021 (4 years 1 month)
Projects (professional or personal) verified_user 0% verified
  • I
    Intrinsic Image Decomposition
    Intrinsic image decomposition separates an image into a reflectance layer and a shading layer. Automatic intrinsic image decomposition remains a significant challenge, particularly for real-world scenes. Implemented the dense CRF-based intrinsic image algorithm for images in the wild and ran experiments with a new prior based on scenery lighting conditions.
  • 1
    1D/2D SLAM
    Implemented Pose Graph Optimisation from scratch using Python. Used EVO & g2o Viewer for evaluating and visualising the trajectory. Also used Bundle Adjustment to perform Monocular vSLAM using sparse matrix representations and generated dense 3D point cloud reconstruction of a world scene made of stereo images from KITTI dataset and recovered the poses using an Iterative Perspective-from-n-points(PnP) algorithm.
This is a community-created genome.