Skip to content
View allenjoshua16's full-sized avatar
🎯
Actively building analytics portfolio
🎯
Actively building analytics portfolio

Block or report allenjoshua16

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
allenjoshua16/README.md

Hi, I'm Allen Joshua 👋

AI Engineer · Agentic Systems · LLM Infrastructure · AI Governance | M.S. Data Science @ UTA

LinkedIn Portfolio Email


🧠 About Me

Building AI systems that don't just work — but work responsibly, reliably, and at scale.

I'm a Master's candidate in Data Science at the University of Texas at Arlington (graduating December 2026), actively exploring the intersection of agentic AI systems, LLM infrastructure, and AI governance through hands-on projects and open source contributions.

When I'm not in class, I'm building — autonomous agents, RAG pipelines, evaluation frameworks, and governance tooling that reflects how I think AI should be deployed: reliably, safely, and accountably.


🚀 What I Build

  • 🤖 Autonomous AI Agents — Multi-step reasoning · Tool use · Memory management · Anthropic Claude API
  • 🔗 Multi-Agent Orchestration — LangChain · LangGraph · MCP Servers · CrewAI
  • 📚 RAG & Knowledge Systems — FAISS · Chroma · Semantic Search · Embeddings
  • 🛡️ AI Governance & Safety — Policy-driven drift detection · Agent trajectory evaluation
  • 🧪 LLM Evaluation Frameworks — Benchmarking · Failure mode analysis · Output auditing
  • Production Systems — FastAPI · Docker · GitHub Actions CI/CD · AWS

🔬 Core Expertise

  • Agentic AI Systems — Building autonomous agents that perceive, reason, and act in production workflows using stateful graph-based systems, tool-augmented agents, multi-agent coordination, and memory management.
  • AI Governance & Policy Enforcement — Contributing to governance tooling with statistical drift checks, policy-driven monitoring, and flexible agent evaluation workflows.
  • RAG & Retrieval Systems — Developing end-to-end RAG pipelines with embedding optimization, chunking strategies, hybrid retrieval, reranking, and vector database management.

📚 What I'm Currently Learning

  • 🔬 Advanced LLM Evaluation — red-teaming techniques, failure mode taxonomy, and building gold-standard ground-truth benchmarks for agentic systems
  • 🏗️ Agent Infrastructure at Scale — distributed agent orchestration, stateful memory management, and multi-agent coordination patterns
  • 🛡️ AI Safety & Governance Engineering — policy enforcement frameworks, model drift detection, bias auditing, and responsible AI deployment practices
  • ⚙️ MLOps for Agentic Systems — experiment tracking, model versioning, canary deployments, and observability tooling for LLM-powered applications

🎯 What Drives Me

I believe the next generation of AI won't be defined by how capable models are — it will be defined by how reliably, safely, and accountably they operate in the real world. That's the problem space I want to work in.


🛠️ Tech Stack

Python Anthropic Claude OpenAI LangChain LangGraph MCP FAISS PyTorch TensorFlow Scikit-learn FastAPI Docker AWS PostgreSQL Redis GitHub Actions


📌 Featured Projects

🔗 Project 📝 Description 🛠️ Stack
🤖 Autonomous Multi-Agent Research System Production-grade agentic AI backend with MCP servers, RAG pipelines, LangGraph workflows, Redis caching, PostgreSQL memory, token cost monitoring, and CI/CD Python · LangChain · LangGraph · Claude API · MCP · FastAPI · Docker · PostgreSQL · Redis · FAISS
📊 AI Data Storytelling Platform 7-stage LLM orchestration pipeline that transforms raw datasets into interactive narratives with deterministic charting and Plotly visualizations Python · Claude API · LangChain · React · FastAPI · Vite · TailwindCSS · Docker
🔬 ViT Robustness Lab Adversarial robustness evaluation framework for Vision Transformers with SHAP attribution, confidence scoring, and interactive failure analysis Python · PyTorch · Vision Transformers · SHAP · Deep Learning · Adversarial ML

🌍 Open Source Contributions

Project Contribution Status
Microsoft Agent Governance Toolkit PR #2858 — Data drift governance checks using KS test, PSI, Chi-squared, and policy-driven enforcement ✅ Submitted
Google ADK Python Issue #4794 — ignore_args support for flexible tool trajectory evaluator comparisons 🔄 In Progress

📄 Publication

📖 Assessment and Prediction of Urban Pollutants and its Influence on Human Health using Deep Learning Algorithm IEEE International Conference on Convergence in Technology (I2CT), 2024


🏆 Recognition

  • 🥇 Most Novel & Creative Award — Adversarial Robustness Framework for Vision Transformers (CSE-5360, Prof. Sihong He, Spring 2026)
  • 📰 IEEE Published Researcher — Deep Learning for Urban Health Prediction, I2CT 2024

📈 GitHub Stats

GitHub Streak


🎓 Education

  • 🎓 M.S. Data Science — University of Texas at Arlington (Jan 2025 – Dec 2026)
  • 🎓 B.E. Information Technology — Don Bosco Institute of Technology, Mumbai University (2020 – 2024)

💬 Let's Connect

I'm actively seeking AI Engineer, Forward Deployed Engineer, ML Engineer, Applied Scientist, AI Governance & Safety Engineering roles for Fall 2026. Open to internships, co-ops, and contract roles. F-1 student with CPT/OPT authorization.

📧 allenjoshua1217@gmail.com | 📍 Arlington, TX | 🌐 Portfolio


Profile Views

Pinned Loading

  1. portfolio portfolio Public

    This is my portfolio — a collection of my data science and analytics experience

    HTML

  2. Tie-Breaker Tie-Breaker Public

    AI-powered strategic decision assistant for structured, real-time, data-driven decision analysis and comparison.

    TypeScript

  3. ViT-Robustness-Lab ViT-Robustness-Lab Public

    An interactive lab for analyzing Vision Transformer robustness through real-time attention map visualization and adversarial attack simulations.

    TypeScript 2

  4. datastory datastory Public

    AI-augmented data storytelling platform that transforms raw datasets into executive-ready narratives, visualizations, and insights using multi-agent LLM orchestration with optional intelligent prep…

    Python