Hi there, I'm Priyank 👋 — aka PriyanK7n
🎓 Graduate Student @ Boston University (ECE + Data Analytics)
💼 Data Engineer @ KGS Technology Group, Inc (Mar 2025 - Present)
🧠 Former AI/ML Dev @ Optimum AI (Microsoft-funded startup)
🌎 Passionate about building scalable AI systems with LLMs, RAG, MLOps, and Cloud infrastructure (AWS | GCP | Azure)
- Deploying AI agents with Crew AI, Mistral LLMs, and private Ollama servers
- Building GitOps-driven CI/CD pipelines using Argo CD, Jenkins, GitHub Actions
- Automating ETL, monitoring, testing, and model retraining pipelines with Airflow, Prometheus, Grafana
- Writing end-to-end data & ML workflows (data ingestion → model training → real-time inference APIs)
💼 KGS Technology Group — Data Engineer (Mar 2025 - Present)
• Building cloud-native pipelines using Airflow, CI/CD, and AWS/GCP
• Driving data governance, quality, and workflow efficiency through cross-functional collaboration
🤖 Optimum AI (Microsoft-Funded) — AI/ML Developer (Apr 2024 – Jun 2024)
• Built a debt negotiation agent using Crew AI + LLM function calling
• Migrated to Ollama inference server with dual Mistral LLMs on Kubernetes (minikube)
• Integrated financial profiling, RAG tools, JSON function calls & deployed on AWS App Runner
📈 Machine Efficiency Prediction via MLOps CI/CD (Jan – Feb 2025)
• Built GitOps-based MLOps pipeline with Jenkins + Argo CD, deployed on Kubernetes (GCP VM)
🛡️ Phishing Detection Pipeline with CI/CD & ML Monitoring (Sep – Oct 2024)
• End-to-end ETL pipeline with MongoDB, feature store, FastAPI, MLFlow, AWS EC2/ECR, Azure ACR
• Integrated logging, data validation, and auto-deployments using GitHub Actions
🛰️ NASA + Titanic ETL Pipelines via Airflow (Jul – Aug 2024)
• Airflow DAGs for ingesting NASA APOD API & Titanic Survival Prediction
• Implemented Redis feature store, Prometheus monitoring, and Grafana alerts
📚 Extractive Summarization with BERT (Nov – Dec 2023)
• Improved BertSum on CNN/DailyMail by 5% using CNN + LSTM + Transformer layers
• Deployed microservices pipeline using Docker, K8s, MLFlow, Prometheus, GCP PostgreSQL
📄 NeurIPS Synthetic Review Dataset + SVD Study (Jul – Aug 2023)
• Built dataset contrasting 1000+ GPT-4 & human reviews; explored SVD recommendation under sparsity variations
🛰️ Point Cloud Reconstruction Using MVS & NERF (May – Jun 2023)
• Reconstructed 3D scenes from 2D images using MVSNet, ZoeDepth, and NERF
🚖 NYC Taxi Demand Analysis on 75GB+ Data (Feb – May 2023)
• ETL & predictive modeling with PySpark, BigQuery, and MLlib, identifying insights across 4 years of taxi trip data
Languages: Python, SQL, Bash
ML/AI: PyTorch, Scikit-learn, Hugging Face, MLlib, LLMs (Mistral, GPT)
MLOps: MLFlow, DVC, Airflow, Prometheus, Grafana, Jenkins, Argo CD
API/Deployment: FastAPI, Streamlit, Flask, Docker, Kubernetes (minikube, K8s)
Cloud: AWS (EC2, ECR, S3, App Runner), GCP (BigQuery, VM), Azure (ACR, Web App)
Databases: PostgreSQL, MongoDB, Redis, GCS
🌐 Blog | ✍️ FastBlogs | 🐦 Twitter
- Analyzing NYC Taxi Trips: Understanding Demand and Optimizing Revenue
- BU EC 503: Recommendation Systems Exploration
- MediBot: Conversational Healthcare Chatbot
- SatFootprint: Building Detection in Satellite Images
- Reformer Reproducibility Challenge @ NeurIPS
💬 “Learn by doing. Stay curious. Build with impact.”
