Shantanu Patne

Algorithms Engineer
AI, Deep Learning, and Signal Processing specialist with industry experience in developing machine learning solutions and real-time systems. Successfully led signal detection systems achieving 95% accuracy and developed facial emotion recognition systems.
View My Work

Professional Experience

Distributed Agile Radio Ecosystems Lab
Graduate Research Assistant
Aug 2024 - May 2025
  • Led research and design of signal detection algorithms using ML models (LSTMs, CNNs, ViT) achieving 95% accuracy over 0dB to 15dB SNR range
  • Pioneered 3 end-to-end pipelines including data generation, augmentation, and custom loss functions with C++ conversion for hardware implementation
  • Estimated signal type and parameters for 10+ signal types using time-frequency analysis
Computational Neuropsychology Lab
Software Developer
Jan 2024 - Aug 2024
  • Built real-time facial emotion recognition system using Residual Attention Networks achieving 92% accuracy
  • Deployed BERT-based NLP prediction system using Docker achieving 88% accuracy
  • Created full-stack training system with Unreal Engine, React, and MongoDB used by 5+ educators and 100+ students
Rakuten Mobile Inc.
Software Engineer
Nov 2020 - Apr 2023
  • Led development of flagship Link RCS application achieving 5M+ users using React, WebRTC, TypeScript, Go, and AWS
  • Improved average app load time by 80% and implemented CircleCI for CI/CD
  • Reduced bug count by 96% improving user satisfaction scores by 50%
  • Collaborated with 5+ cross-functional teams to implement 15+ new features

Technical Skills

💻

Programming

Python C/C++ MATLAB JavaScript CUDA TypeScript Go
🤖

AI/ML Frameworks

PyTorch JAX TensorFlow OpenCV TinyML ELL
🧠

AI/ML Models

CNN LSTM ViT BERT NeRF LLaMa Wav2Vec
🔧

Embedded Systems

STM32 Jetson Raspberry Pi ESP32 FreeRTOS Docker

Featured Projects

Hybrid Inpainting of 3D Meshes
Jan 2025 - May 2025

Researched mesh inpainting techniques comparing 3+ traditional geometric and generative approaches. Achieved 26% improvement in Haussdorff distance and 250% PSNR increase using NeRF-based approaches.

Improving Wav2Vec 2.0 with Wave-U-Net
Aug 2024 - Dec 2024

Developed noise-robust speech recognition pipeline achieving 68% WER reduction (from 99.77% to 31.69%) at -10 dB SNR conditions by combining Wave-U-Net source separation with Wav2vec2.0.

Fractals Dynamics in NN Boundaries
Jan 2024 - Jun 2024

Investigated fractal boundaries in neural network trainability using chaos theory analysis and maximum Lyapunov Characteristic Exponents (mLCE) across varying architectures with fractal dimension of 1.3429.

Education

MS, Electrical Engineering
Arizona State University
GPA: 4.0/4.0 | May 2025

Key Coursework:

Time-Frequency Analysis, Real-time DSP, Multidimensional Signal Processing, Speech Recognition & Compression, Mathematical Foundations of ML, Machine Vision & Pattern Recognition, Spatial Audio, IoT, XR Design, Sensor Fusion

B.E, Electronics Engineering
Pune University
GPA: 9.07/10 | Aug 2020