Amith Penumudi

About Amith Penumudi

I am a results-driven professional with a strong foundation in software engineering, having built reliable, scalable, and production-ready systems. My experience includes developing robust backend and full-stack tools, improving system stability, optimizing workflows, and addressing InfoSec challenges in fast-paced, enterprise environments.Building on this foundation, I have developed expertise in data science and machine learning. I have worked on predictive modeling, deep learning, and transformer-based architectures, applying them to real-world problems such as weather forecasting and financial time series analysis. My experience spans the complete ML workflow—data preprocessing, feature engineering, model training, evaluation, and deployment—while leveraging software engineering best practices to ensure scalable and maintainable solutions.Key Skills & Technologies• Python | Java | C++| C | JavaScript | SQL • Machine Learning | Deep Learning | Transformers | Time Series | Data Analysis & Visualization | AI• Docker | Kubernetes | Microservices |• CI/CD | GitHub Actions | Datadog | DBMSImpact & Experience Highlights• Developed scalable call-processing and observability tools at Uniphore, improving system reliability, stability, and monitoring while resolving critical InfoSec issues.• Built and evaluated machine learning and deep learning models at Jio to predict temperature and heatwave occurrence, leveraging regression, classification, and feature engineering on large-scale climate data.• Built a GPT-style Transformer from scratch (11M parameters) to generate Shakespeare-style text, implementing multi-head attention, positional embeddings, and an optimized GPU-accelerated training pipeline.• Built ML/DL models to classify swing points in stock prices, engineered features from price movements, handled class imbalance, and post-processed predictions for accurate results.Feel free to reach out if you’d like to discuss data analytics, data science, machine learning, ai or potential collaboration: a••••••••@gmail.com

Experience

1. Software Engineer

   Uniphore

   Jun 2024 — Present · Chennai, IN

   Developed a call-processing tool to replicate agent-side operations in a call centre environment, enabling smooth handling of calls with features like hold, transfer, and conference.• Enhanced stability and ensured zero downtime across multiple services through careful code refactoring, stability improvements, and proactive monitoring.• Implemented support for multiple token types, extracting required data dynamically based on the token received.• Designed and published metrics to Datadog, tracking requests and responses across multiple services for enhanced monitoring and observability.• Worked on multiple InfoSec issues, identifying and mitigating security vulnerabilities to strengthen system integrity and compliance.• Collaborated with cross-functional teams to optimise workflows, improve code maintainability, and deliver high-quality software solutions at scale.

Education

• National Institute of Technology Karnataka

  Bachelor's degree