Skip to content
View JoshPola96's full-sized avatar

Block or report JoshPola96

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 supported. This note will be visible to only you.
Report abuse

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

Report abuse
JoshPola96/README.md

Hi 👋, I’m Joshua Peter Polaprayil

🚀 **AI Engineer · Full-Stack Developer · MLOps Enthusiast** 🌍 Open to work globally · 💼 Open to freelance / contract / full-time roles · 🤝 Open to collaboration & mentorship

🧠 About Me

I am a results-driven AI/ML Engineer with an MSc in Big Data Analytics & AI, focused on bridging the gap between research models and production systems. I don't just build models; I build the API, the infrastructure, and the deployment pipelines to make them work in the real world.

Most recently, I operated as a Freelance AI Engineer for EsimTime, where I solo-architected and deployed a fully autonomous customer support agent handling global eSIM commerce. I am now looking to bring this hands-on production experience to a full-time engineering team or independent role where I can contribute immediately and grow my career.


💼 What I Bring to a Team

  • Production Engineering: I build systems with Docker, CI/CD (GitHub Actions), and robust error handling—not just Jupyter notebooks.
  • Modern Agentic AI: Deep experience with LangGraph, Gemini Flash 2.0, and RAG architectures.
  • Full-Stack Integration: I connect AI to the real world using FastAPI, PostgreSQL, and Redis.
  • Global Mindset: Experience working with international clients and delivering enterprise-grade documentation.

🧭 Project Highlights

  • Freelance AI Engineer — EsimTime (2024)

    • The Challenge: Automate customer support for a global eSIM provider across Telegram & WhatsApp.
    • The Solution: Built a production-grade autonomous agent using LangGraph, Qdrant, and Gemini 2.0.
    • Key Achievement: Achieved <3s latency for complex queries and implemented identity-aware authentication without human intervention.
    • 👉 View the Architecture Case Study
  • Enterprise‑Scale MLOps Project (Open Source) Built a self‑healing bankruptcy prediction system with automated training, monitoring, and retraining pipelines.

  • AI Intern — Computer Vision & ADAS Customized YOLO‑based models, engineered training optimizations, and built API‑ready ML pipelines for real‑time insights.

  • Junior Software Engineer — Financial Systems Contributed to enterprise loan decision engines using C#/.NET with SQL Server, debugging production workflows and supporting live deployments.


🧰 Tech Stack

🤖 AI / Machine Learning

Python · PyTorch · TensorFlow · scikit‑learn · Transformers (BERT, BioBERT) · LangChain · LangGraph · RAG · NLP · Computer Vision (YOLOv8, OpenCV, OCR)

🧠 Data & Analytics

Pandas · NumPy · PySpark · Databricks · Feature Engineering · Clustering · Regression · Data Visualization

⚙️ Backend & APIs

FastAPI · Flask · NestJS · .NET Core · REST APIs · JWT Auth · SQLAlchemy · Prisma

🧩 MLOps & Infrastructure

Docker · Docker Compose · GitHub Actions · CI/CD · Model Monitoring · Retraining Pipelines · Celery

☁️ Cloud & DevOps

AWS · Google Cloud Platform (GCP) · Linux · Secure Deployments · Environment Management

🖥️ Frontend & Apps

Streamlit · React · TypeScript · HTML · CSS · Bootstrap

🗄️ Databases

PostgreSQL · SQLite · SQL Server · Vector Databases


📌 Featured Work

My pinned repositories highlight industry‑grade systems, including:

  • Agentic RAG chatbots — multi-agent AI assistants with persistence and knowledge integration
  • Enterprise MLOps pipelines — scalable, automated AI workflows
  • Computer vision & deep learning systems — real-time detection, tracking, and predictive modeling
  • Full-stack AI applications — production-grade systems with API, backend, and frontend integration
  • Each repository reflects real design decisions, trade‑offs, and production considerations — not toy demos.

🌍 Find Me Elsewhere

📄 Resume: (PDF link coming soon) 💼 LinkedIn: https://www.linkedin.com/in/josh33-peter10/ 🐙 GitHub: https://github.com/JoshPola96

GitHub followers GitHub stars


⭐ If you’re a recruiter, collaborator, or founder — feel free to explore my repos or reach out. I’m always happy to talk about systems, trade‑offs, and ideas.

Pinned Loading

  1. enterprise-ai-agent-architecture-esim enterprise-ai-agent-architecture-esim Public

    Enterprise AI Agent Architecture (EsimTime): A production-grade autonomous agent case study handling global eSIM support. Features a full-stack architecture with LangGraph, Gemini Flash 2.0, RAG, a…

  2. company-bankruptcy-prediction-mlops company-bankruptcy-prediction-mlops Public

    Enterprise-grade MLOps pipeline for predicting company bankruptcy. Automates data ingestion, model training, monitoring, retraining, and deployment on AWS with Airflow, MLflow, Terraform, and Strea…

    Jupyter Notebook 3 3

  3. ai-assistant-hub-langgraph ai-assistant-hub-langgraph Public

    A modular, multi-agent AI assistant framework orchestrating specialized agents for RAG-based QA, translation, real estate insights, document summarization, and utility tasks. Built with Streamlit a…

    Python

  4. medical-chatbot-biobert-transformer medical-chatbot-biobert-transformer Public

    Research-grade medical AI chatbot combining BioBERT embeddings with a custom multi-layer Transformer decoder. Provides interactive Q&A for doctor–patient style queries via Streamlit, supporting Bea…

    Python

  5. dissertation-racism-detection-bert-cnn-bilstm dissertation-racism-detection-bert-cnn-bilstm Public

    Bias-minimized multimodal deep learning system for racism detection in social media, combining BERT-based NLP with structured feature modeling as part of an MSc dissertation.

    Jupyter Notebook 3 2

  6. heart-attack-data-pipeline heart-attack-data-pipeline Public

    End-to-end data engineering pipeline for analyzing heart attack prediction in Indonesia. Automates data ingestion, transformation, and visualization using GCP (Terraform, BigQuery, Cloud Storage), …

    Shell