Hello! 👋

I'm Sushant Gautam

Software Engineer | MLOps | Gen AI

About Me

Software Engineer with a passion for building intelligent and generative AI systems that scale. Experienced in LLMs, RAG, LangChain, and MLOps (Docker, Kubernetes, CI/CD), with deep expertise in computer vision, deep learning, and cloud deployments. I enjoy architecting robust data pipelines, optimizing models, and delivering reliable AI products that solve real-world problems. Open to full-time roles in software engineering, AI/ML or MLOps.

Python Java C++ CUDA AutoML TypeScript Next.js FastAPI SQL PyTorch TensorFlow Scikit-learn Computer Vision GANs Diffusion Models LLMs GenAI RAG Agentic AI LangChain LLMOps Hugging Face MLOps Docker Kubernetes MLflow DVC Airflow Kafka PySpark AWS GCP CI/CD Jenkins Prometheus Grafana Cassandra ROS2 Git Linux Data Engineering

Projects

01

📈🤖 AI Stock Advisor

Streamlit-powered app for real-time stock analysis and company research. Integrates Yahoo Finance data, LLM agents, and Gemini for personalized investment reports and actionable insights.

PythonStreamlitLLMsGeminiAgentic AIyfinanceData EngineeringPrompt Engineering
02

🧩⚡ TabGist — Chrome Summarizer Extension

Chrome MV3 extension with FastAPI backend for summarizing web pages and YouTube transcripts into concise, print-ready notes. Features local-first processing, Gemini-powered output, and multi-language support.

JavaScriptChrome ExtensionFastAPIPythonLLMsGeminiNLPREST APIs
03

🩺 Medical RAG Chatbot

End-to-end Retrieval-Augmented Generation chatbot for medical Q&A. Uses LLMs, LangChain, and Pinecone for vector search over curated medical data, with Flask UI and AWS CI/CD deployment. Delivers context-grounded, reliable answers.

PythonLLMsRAGLangChainPineconeFlaskVector SearchEmbeddingsAWSCI/CD
04

End-to-End MLOps: Image Classification

Robust ML pipeline for image classification: data ingestion, training, evaluation, and automated CI/CD deployment on AWS SageMaker. Includes MLflow tracking and DVC for reproducibility.

PythonPyTorchSageMakerMLflowDVCAWSMLOpsCI/CDData Engineering
05

Real-Time Data Streaming Pipelines

Scalable pipelines for real-time data ingestion, transformation, and serving. Built with Airflow orchestration, Kafka streaming, PySpark processing, Docker containers, and Cassandra storage.

PythonAirflowKafkaPySparkDockerCassandraData EngineeringStreaming
06

Self-Driving Car (End-to-End DL)

Autonomous driving stack for steering angle prediction using deep learning on front-camera frames. Includes data preprocessing, model training, and real-time inference with OpenCV.

PythonPyTorchComputer VisionOpenCVDeep LearningAutonomous Vehicles
07

Handwritten Line Text Recognition (OCR)

Handwritten text recognition using CRNN (CNN+LSTM+CTC) architecture. Achieved low CER on IAM dataset with advanced data augmentation and real-time demo interface.

PythonTensorFlowOCRComputer VisionDeep LearningData Augmentation
08

ROS2 Autonomous Tractor

Autonomous tractor navigation using sensor fusion (LIDAR, GPS, RGB-D) and ROS2. Developed control stack for real-time perception, localization, and path planning in outdoor environments.

PythonROS2Sensor FusionPerceptionControlsLocalizationPath Planning

Experience

SDE Intern

Amazon (AWS) — Transcribe Streaming

May 2025 – Jul 2025
  • Built a Next.js/TypeScript dashboard summarizing 1,000s of integration test runs in <5s.
  • Zero-touch deployments with AWS CDK + Elastic Beanstalk; prod rollout in <10 min.
  • Failure summarization reduced root-cause analysis time by ~40–60%.

Graduate Research Assistant (CV/GenAI)

Mississippi State University

Aug 2023 – May 2025
  • Image restoration research (GANs, diffusion) with Stanford/Princeton/BlueHalo; +20% quality.
  • CI/CD for ML (Docker, K8s, Airflow); training/eval with PyTorch-Lightning, Optuna.

Machine Learning Engineer

Proxmed Pty. Ltd.

Feb 2023 – Aug 2023
  • LVO detection with YOLOv7; +8% accuracy; serverless deploy via AWS Lambda.
  • Model optimization (quantization/ONNX) and CI/CD (GitLab, Docker, K8s).

ML/CV Engineer & AI Research Scientist (Contract)

Wiseyak / HeHealth (prior roles)

Oct 2020 – Apr 2022
  • Medical imaging models; reduced false positives and improved diagnostic accuracy.
  • Synthetic data (GANs) + few-shot learning boosted model accuracy.

Education

M.S. in Computer Science (AI), GPA 4.0

Mississippi State University

Aug 2023 – Aug 2025
  • Coursework: Algorithms, Cloud Computing, AI Robotics, Autonomous Vehicles, Data Science, Deep Learning
  • Thesis: X-DECODE — EXtreme Deblurring with Curriculum Optimization and Domain Equalization (arXiv, 2025). https://arxiv.org/abs/2504.08072
  • President, CSE AI Club; organized GenAI workshops and hackathons