FlowCV Logo
Samanvya TripathiAI & ML Engineer
Education
Masters in Computer Science, Northeastern University
Graduating Dec 2024 | Boston, MA

Relevant Coursework: Machine Learning (ML), ML Operations, Natural Language Processing (NLP), Algorithms, Programming Design Paradigm, Databases

Skills
Machine Learning: Python, PyTorch, Tensorflow, Scikit-Learn, XGBoost, TensorRT, TFLite, CUDA, Joblib, Dask, GPT, BERT, Transformers, GAN, VAE, RAG, Llamaindex, Huggingface, Langchain, Paddle, Optuna, MLFlow, Raytune
Software Engineering: JavaScript, Java, Design Patterns, TypeScript, NodeJS, Express, React, MongoDB, MySQL, Docker, AWS, Azure, Git, CI/CD, Github, FastAPI, Flask,
Awards
Hack UMass 2022, University of Massachusetts Amherst and MLH
Nov 2022

Hackthon Winner: Best AI/ML Hack. Project TItle: Meta-Identity

Hack Harvard 2022, Harvard University and MLH
Oct 2022

Hackathon Winner: Best Use of Google Cloud (GCP). Project Title: ReAlive

Work Experience
Software Engineer Co-Op, Honeywell
Jan 2024 – Aug 2024 | Fort Mill, SC
  • •Developed transformer-based image enhancement models in TensorFlow, using datasets for a 32% KPI improvement in low-light conditions.
  • •Led the integration of computer vision models into production for a November 2024 launch, ensuring reliable deployment on edge devices.
  • •Streamlined the model lifecycle with MLflow, enhancing tracking, versioning, and deployment on Azure, reducing time-to-deployment by 20%.
  • •Converted and optimized models to ONNX and TFLite, applying selective quantization for improved accuracy and latency on NXP edge devices.
  • •Managed data preprocessing and built pipelines for 100,000+ images, ensuring efficient data flow and enabling high-quality training.
    • Graduate Teaching Assistant, Khoury College of Computer Sciences
    Sep 2023 – Dec 2023 | Boston, MA
  • •Guided 200+ students on advanced ML concepts using Python, Pandas, and Scikit-Learn, improving problem-solving and code optimization.
    • Machine Learning Engineer, EdgeNeural AI
    Jan 2022 – Jun 2022 | Pune, India
  • •Utilized neural architecture search (NAS) to optimize object detection models, improving deployment efficiency and performance metrics.
  • •Created an API with FastAPI and Docker to optimize deep learning models, cutting server load by 30% and boosting response time.
  • •Integrated AWS Sagemaker Neo service, expanding deep learning framework compatibility and edge device support by 42%.
    • Machine Learning Engineer, Myelin Foundry
    Jun 2020 – Sep 2021 | Bengaluru, India
  • •Developed Deep Learning models for Super-Resolution and Denoising on edge devices, boosting image quality by 10% and FPS by 20%.
  • •Established an efficient MLOps pipeline with Azure, CI/CD and Docker, reducing model development timelines by 40%.
  • •Leveraged 1,000+ videos for model training, utilizing parallel data processing and optimization techniques for pre- and in-training efficiency.
    • Machine Learning Intern, Myelin Foundry
    Jan 2020 – Jun 2020 | Bengaluru, India
  • •Analyzed data strategies to enhance SR products for OTT platforms on Xilinx edge devices, optimizing FPGA performance.
  • •Engineered a Docker environment for ML applications, streamlining development and cutting setup time by 50%.
  • •Improved OTT video streaming by fine-tuning encoder settings, enhancing bitrate efficiency by 15% using FFmpeg and TensorFlow.
    • Projects
    Meta-Identity

    Developed AI-generated metaverse avatars using GPT-2, GAN, and StyleGAN, enhancing character realism and interactivity.

    NASDAQ Stock Price Prediction

    Developed predictive models for Nasdaq closing cross auction prices using CatBoost, Linear Regression, and Neural Networks, enhancing time series analysis and accuracy.

    Lyrics Generator

    Designed a generative AI model using LSTM, BERT, and GPT architectures to create original song lyrics for artists like Taylor Swift, demonstrating natural language processing capabilities.

    GTA V Player Bot

    Built an AI-powered bot capable of navigating complex 3D game environments like GTA V and Witcher 3 using advanced computer vision techniques to detect, interpret, and interact with virtual scenes.