Riddhima Medatwal
Building scalable systems, intelligent models, and autonomous solutions. Passionate about creating production-ready software that solves real-world problems across multiple domains.
Experience
Building software, systems, and solutions across multiple domains
Accenture
Working on NLP-based systems for misinformation detection and developing transformer architectures for classification tasks.
- • Designed and implemented an NLP-based classification system for detecting misinformation, processing 170,000+ news articles
- • Developed transformer architectures (BERT, RoBERTa), achieving ~20% improvement in validation accuracy over initial benchmarks
- • Optimized preprocessing and feature extraction pipelines in Python (Pandas, scikit-learn) for faster testing with neural networks
PRACSYS Lab
Researching deep learning approaches for humanoid robot motion prediction, developing CNN architectures for trajectory forecasting.
- • Conducting research on probabilistic CNN-based humanoid robot trajectory prediction in collaboration with doctoral researchers
- • Developed a Convolutional Neural Network trajectory prediction model, improving accuracy compared to other architectures
- • Processed and analyzed large-scale datasets for predictive modeling, implementing normalization and feature extraction pipelines
Cornell University - Break Through Tech
Selected for intensive machine learning program, leading projects applying ML and deep learning to real-world problems.
- • Chosen from over 3,000+ applicants for an intensive summer machine learning curriculum taught by Cornell University faculty
- • Spearheaded projects applying machine learning algorithms and deep learning frameworks to real-world problems
- • Applying AI principles and tools such as Scikit-Learn, Pandas, Numpy, PyTorch, TensorFlow, and more
Rutgers VEX Robotics Team
Leading software development for competitive robotics, engineering autonomous control algorithms and optimizing robot performance.
- • Achieved 1st place in New Jersey and qualified for the VEX Robotics Worlds Championship 2025
- • Engineered and optimized autonomous control algorithms that improved robot performance efficiency by 32%
- • Spearheaded real-time sensor-based movement and optimization efforts to reduce latency and improve control consistency
DATA STORTELLING LLC
Developed off-grid wireless communication systems for rural emergency communications, integrating hardware and software components.
- • Designed and built custom LoRa-based mesh nodes using Meshtastic for fully offline communication, useful in emergency scenarios
- • Built an end-to-end hardware system with the integration of GPS, OLED, antenna components to create reliable, user-friendly devices
- • Implemented and tested mesh networking protocols for resilient communication in low-connectivity environments
Skills & Expertise
Technical skills across software engineering, machine learning, and robotics
Software Engineering
Machine Learning & AI
Robotics & Systems
Machine Learning & AI
Machine Learning & AI
Developing NLP systems for real-world applications, working with transformer architectures and large-scale datasets. Experience with natural language processing, computer vision, and building production-ready ML models.
Technical Skills
NLP & Transformer Models
Building transformer-based classification systems for misinformation detection, processing large-scale text datasets and optimizing model performance.
Robotics & Autonomous Systems
Robotics & Autonomous Systems
Building autonomous systems across competitive robotics and research applications. Leading software development for competitive teams while conducting research on humanoid robot trajectory prediction. Experience with sensor integration, real-time control, autonomous algorithms, and optimizing robot performance.
Technical Skills
Autonomous Control & Optimization
Engineering autonomous control algorithms for competitive robotics, achieving 32% efficiency improvement through real-time sensor-based optimization.
Featured Projects
Software engineering, ML/AI, and robotics projects
Claryn: An Autonomous Machine Learning Agent
Designed and built an autonomous AutoML agent that cleans datasets, performs multi-model training and evaluation, and automatically compiles enhanced models, metrics, and Jupyter notebooks for streamlined reproducibility and comparison. Implemented model optimization feature that ingests additional datasets to improve existing models and delivers updated models with comprehensive improvement metrics.
FreshShelf
Developed an iOS application that utilizes real-time speech recognition for voice-based entry leveraging Core Data to log food items and expiration dates, enabling fully hands-free entry and reducing manual user interaction by over 70%. Designed and deployed expiration alerts that show upcoming expirations with personalized notification timing.
mock-server-proxy
Built and published mock-server-proxy (npm), an open-source CLI tool serving parameterized JSON mocks with latency injection, hot reload, and endpoint versioning to accelerate full-stack development workflows. Added proxy and record modes to capture, store, and replay live API calls as mocks, cutting integration debugging time by 50%.
Let's Connect
Interested in new opportunities, collaborations, and conversations about technology and design.
riddhima.medatwal@gmail.com