Skip to content
View Jeet-51's full-sized avatar

Highlights

  • Pro

Block or report Jeet-51

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Jeet-51/Readme.md

About Me

class Jeet:
    role     = "Software Engineer | AI/ML Engineer"
    degree   = "M.S. Data Science Β· Indiana University Β· GPA 3.8/4.0"
    location = "Bloomington, IN β†’ Open to Relocation"

    focus = [
        "Production LLM systems & GPU inference optimization",
        "High-throughput backend services with AI/ML integration",
        "Distributed data pipelines & MLOps at scale",
        "Reliable, observable, maintainable software",
    ]

    currently_building = "LLM inference APIs + backend systems that work in prod, not just notebooks"

Impact at a Glance

πŸš€ Metric
14M+ Nonprofit records served in production
1,600Γ— Latency reduction via Redis caching (8.3s β†’ 5ms)
180ms P50 inference latency on GPU-backed LLM service
500K+ Creator–brand interactions on AI-powered platform
0.95 RΒ² on Medicare billing prediction pipeline
78 Latent funding networks discovered via Neo4j

Featured Projects

⚑ LLM Inference Service

Production-grade API with 1,600Γ— latency reduction

  • vLLM + AWQ quantization β€” 28GB β†’ 14GB VRAM
  • Redis semantic cache β€” 8.3s β†’ 5ms response time
  • 180ms P50 across 100+ concurrent requests
  • Prometheus + Grafana observability stack

vLLM FastAPI Redis CUDA Docker Prometheus

πŸ“‚ Repo

πŸ”„ Subscription Commerce Backend

Stripe-style payment backend in Go

  • Idempotency middleware β€” zero duplicate charges
  • 20ms P50 latency Β· 0% error rate in load testing
  • ACID transactions with full rollback guarantees
  • Redis rate limiting per API key

Go PostgreSQL Redis Docker

πŸ“‚ Repo Β· πŸ“ Blog

🎯 PitchPal β€” AI Startup Evaluator

LangChain ReAct agents for due diligence

  • 60% faster investor due diligence cycle
  • Multi-agent architecture with real-time market analysis
  • Processed 100+ startup pitch evaluations

LangChain GPT-4 Streamlit Python

πŸ“‚ Repo Β· 🌐 Live Demo

πŸ›‘οΈ ClaimGuard β€” Healthcare ML Pipeline

End-to-end Medicare billing prediction system

  • RΒ² = 0.95 on unseen billing data
  • 200K+ claims with Bio_ClinicalBERT embeddings
  • MLflow tracking + SHAP explainability
  • Delta Lake versioned feature store

XGBoost PySpark MLflow SHAP Transformers

πŸ“‚ Repo

πŸ’° FinanceFlow β€” Fraud Detection Pipeline

Real-time cloud ETL with ML-powered detection

  • +40% fraud detection accuracy
  • S3 β†’ PySpark β†’ SageMaker real-time pipeline
  • Automated alerting on anomaly patterns

AWS PySpark SageMaker Redshift Lambda

πŸ“‚ Repo

🌍 EPA GHG Analytics Platform

Production analytics stack for climate data

  • 40% less analysis time for data teams
  • Snowflake + dbt modular transformation pipeline
  • Interactive Tableau dashboards with scenario modeling

Snowflake dbt Tableau Python SQL

πŸ“‚ Repo


Tech Stack

Languages Python Go Java SQL

AI / ML PyTorch TensorFlow LangChain HuggingFace scikit-learn OpenAI

Backend & Infra FastAPI Docker Kubernetes Redis PostgreSQL AWS

MLOps & Data MLflow Airflow Spark Snowflake dbt Neo4j


GitHub Stats


Latest Writing

πŸ“ I Built a Subscription Backend Like Stripe in 6 Hours β€” Here's What I Learned


Let's Connect

I'm open to full-time roles in software engineering, AI/ML engineering, and backend systems.

Pinned Loading

  1. PitchPal-AI-Agent-for-Startup-Pitch-Evaluation PitchPal-AI-Agent-for-Startup-Pitch-Evaluation Public

    A production-grade AI agent that evaluates startup pitches using a custom ReAct (Reasoning + Acting) loop

    Python

  2. ClaimGuard-Intelligent-Healthcare-Service-Pattern-Analysis ClaimGuard-Intelligent-Healthcare-Service-Pattern-Analysis Public

    A scalable pipeline for intelligent Medicare claims analysis using NLP, Delta Lake, and MLOps.

    Jupyter Notebook

  3. FinanceFlow-Cloud-Enabled-Analytics-for-Fraud-Detection FinanceFlow-Cloud-Enabled-Analytics-for-Fraud-Detection Public

    Python

  4. Humanizer-Bot Humanizer-Bot Public

    AI Humanizer

    TypeScript 1 1

  5. Prediction-System-Design-for-Monitoring-the-Health-of-Developing-Infants-using-Statistical-ML Prediction-System-Design-for-Monitoring-the-Health-of-Developing-Infants-using-Statistical-ML Public

    Segregation models aid healthcare diagnosis, reducing errors. Extracting real-world medical data is challenging due to variability. This study automates distinguishing cardiotocographic regions in …

    Jupyter Notebook

  6. EPA-Greenhouse-Gas-GHG-Emissions-Dashboard EPA-Greenhouse-Gas-GHG-Emissions-Dashboard Public

    Developed an EPA GHG Emissions Dashboard (2021–2023) using Snowflake, dbt, SQL, and Tableau. Built star-schema marts and KPI tables to cut query latency by 50%. Applied advanced Tableau techniques …