Hemanth Vikash Kannan Rajan

+1-540-***-****, adzf43@r.postjobfree.com Brooklyn NYC - 11226 LinkedIn

WORK EXPERIENCE

Machine Learning Engineer, Moshman Research May 2021 – Present

• Spearheaded the development of machine learning solutions by researching academic papers with open-sourced code, resulting in a remarkable 10% increase in project proposals successfully submitted to government agencies.

• Constructed custom Physics informed Neural Network (PINN) model that is time dependent (as a variation of RNNs) with a dataset made of 50 windmill climbers, attaining an accuracy level of 86%.

• Trained 50+ model iterations to maximize efficiency and accuracy of the XGBoost algorithm results as a proof of concept, achieving prediction success rate of up to 97%.

• Conducted quantitative market trend research and analyses of competitor product offerings, enabling the successful deployment of personalized marketing strategies for a potential client resulting in $300K additional revenue.

• Coordinated development of a $5.5M data visualization system to improve Space Force’s Commander’s dashboard, resulting in 40% more efficient decision making & shorter processing times.

• Integrated data streams with the Commander’s Dashboard to create an effective real-time analytics system, supporting up to 50 active users and about 5 custom integrations.

• Transformed complex datasets into interactive data visualization blueprints, consolidating key learnings to facilitate decision-making with an 75% reduction in development time. Lead Data Scientist, Gander Aug 2020 – May 2021

• Achieved 10x faster training speeds by optimizing model structures and refactoring codebase across 600k data points, resulting in a 12% improvement in precision.

• Coordinated efforts of a 4-person team to construct a novel skincare recommendation engine using state-of-the-art PyTorch and sklearn algorithms, meeting all client requirements.

• Earned shareholder support for the 6 new data acquisition methods through a comprehensive analysis of potential market sustainability, resulting in an 18% increase in ROI.

• Constructed a custom K-Means clustering algorithm to accurately categorize Gander's clientele according to hair & skin types, resulting in increased sales conversion of 10%. PROJECTS

REMind: iOS data-driven Application Aug 2018 – May 2022

• Developed an innovative iOS application on SwiftUI for sleep analysis built to access Apple Watch sleep data

• Integrated Open.ai’s GPT-2 & GPT-3 models to improve NLP tasks such as text generation & emotion detection with 15% reduction in false positives.

• Integrated Firebase with Tensorflow to mine & analyze over 5K hours of sleep data from 10 test users, resulting in a 80% accuracy in similarities predictions between dream descriptions & sleep data. False News and Clickbait Detector Jan 2020 – May 2020

• Developed an ML algorithmic solution with 94% accuracy to classify & filter verified news articles from false reports.

• Constructed custom classifier model using NLP techniques, improving existing accuracy rate in the discipline by 6%.

• Engineered 22 features leveraging NLTK in Python to enable automated text analytics functionalities, achieving 20X faster performance for data extraction.

RESEARCH EXPERIENCE

Machine Learning Researcher, Unmanned Systems Lab, Virginia Tech June 2019 – Aug 2019

• Validated YOLO suite of algorithms’ performance through comprehensive analytics, total false positives down 15%

& probability of correct detections up 10%.

• Optimized web-scraping framework to scrape 75K images from Google, reducing manual collection time by 80%.

• Wrote Python script to synthesize COCO dataset of 200K images, reducing annotation time by 10X. Data Analysis Researcher, Cyber-Human Systems, Virginia Tech Aug 2018 – May 2019

• Configured survey infrastructure using SQL and Python backend to capture 2K responses/hour for survey analysis

• Integrated cutting-edge NLP algorithms and machine learning techniques into hate speech prediction model design, increasing recognition rate of highest priority cases by 10%. EDUCATION

Bachelor of Science, Computer Science with Machine Learning Capstone Virginia Tech CERTIFICATES

• Convolutional Neural Networks

DeepLearning.AI, June 2022

Skills: Building CNNs from scratch

• Sequence Models

DeepLearning.AI, Aug 2023

Skills: Transformer Networks, RNNs,

Attention Models, Word Embeddings

• Neural Networks and Deep Learning

DeepLearning.AI, Feb 2022

Skills: Building Neural Networks in

Python

TECHNICAL SKILLS

Machine Learning tools: Sci-kit, Scipy, Keras, numpy, pandas, Apache Spark, Jupyter Notebook, Plotly, Seaborn, H2O

Deep Learning: Transformer Models, GAN, Sequential Models, PINNs Statistical analysis: Python, MATLAB, basic knowledge in R Stack Programming: Java, Python, Objective C, Swift, C/C++ Database Management: SQL, MongoDB

Contact this candidate