AI/ML Systems Engineer · Backend & Data Platform Builder
Designing secure, production-grade ML systems, RAG platforms, and data-intensive services.

• 🔭 Current focus: AI governance frameworks, enterprise AI agents, and ML systems at scale
• 🧠 Experience: 4+ years across ML, backend (FastAPI), and data engineering / MLOps
• 🌍 Location: Dhaka, Bangladesh — open to global & remote collaboration
• 🧭 Interest areas: AI governance, RAG platforms, ML infra, and long-term impacts of AI

🧮 Machine Learning & LLMs

• Supervised / unsupervised ML, classic models to modern deep learning
• Retrieval-Augmented Generation (RAG), semantic search, vector databases
• Generative AI (Llama 2, Gemma, AWS Bedrock) and evaluation / observability
• Responsible AI and AI governance design for real-world organizations

🧵 Backend & APIs

• FastAPI-based microservices and high-performance REST APIs
• Async Python, streaming responses, background jobs, task queues
• Auth (JWT), rate limiting, observability, and clean modular architecture
• Integration with databases, vector stores, and external ML services

📊 Data Engineering & MLOps

• ETL/ELT pipelines, data modeling, and analytics-ready warehouse design
• CI/CD for ML, Docker-first workflows, cloud deployment (AWS/GCP)
• Experiment tracking (MLflow / W&B / Neptune), model versioning
• Production monitoring, logging, and feedback loops for ML services

• Languages: Python, SQL, Bash, JavaScript (frontend basics)
• ML / Data: scikit-learn, pandas, NumPy, PyTorch / TensorFlow (as needed), Hugging Face, LangChain, vector DBs (Qdrant, PGVector)
• Backend: FastAPI, REST APIs, async Python, JWT auth
• Data & Storage: PostgreSQL, MongoDB, relational modeling, query optimization
• MLOps / Infra: Docker, Airflow, GitHub Actions / GitLab CI, AWS, GCP, MLflow, W&B, Neptune
• Analytics & BI: Power BI, Tableau, Superset
• Tools: Git & GitHub, CLI, Postman/curl, VS Code, Jupyter

• 🧩 Production ML Systems – taking models from notebooks to robust, observable services
• 📚 RAG & Search – domain-specific assistants with strong retrieval, ranking, and evaluation
• 🛡️ AI Governance – policies, frameworks, and tooling for secure, compliant, responsible AI adoption
• 🧰 MLOps – making training, deployment, and monitoring repeatable, automated, and auditable

These are good candidates to pin on GitHub.

• LLM Data Analyzer (Llama 2 + Streamlit + Docker)
  Streamlit-based app powered by a Dockerized Llama 2 model that provides conversational insights and statistical summaries from CSV/Excel datasets.

• RAG Observability Platform (Docker + MLOps)
  A Dockerized observability stack for RAG pipelines with real-time tracing, quality metrics, and evaluation hooks (ideal for production RAG systems).

• AI Incident Reporting Agent (NLP + RAG + PGVector)
  Enterprise incident reporting assistant that automates categorization and summarization of reports, reducing manual processing by ~90% in production settings.

• Privacy-First Enterprise RAG System (FastAPI + Qdrant)
  A domain-specific semantic search and question-answering system designed with data isolation and governance in mind, suitable for compliance-sensitive environments.

• MLOps Pipeline for CIFAR-10 (Docker + CI/CD)
  End-to-end image classification pipeline with experiment tracking and production-ready deployment flow, achieving 98% validation accuracy.

• ETL Pipeline with Airflow & Docker
  Orchestrated ETL that processes 100K+ records daily and reduces data refresh times from hours to minutes.

• End-to-end ML projects with:
  • Clear READMEs, problem statements, and architecture diagrams/overviews
  • Reproducible environments (Docker, requirements, Makefiles or helper scripts)
  • CI/CD scaffolding and monitoring hooks where relevant
• Backend services built with FastAPI and solid API design principles
• Data pipelines and utilities for ETL, feature engineering, and analytics
• Experiments with LLMs, vector search, and RAG patterns (incl. observability & evaluation)
• Occasional open-source contributions and starter templates you can reuse in your own projects

Recommended pinned repos:

• ✅ A production-style ML service (FastAPI API + model + monitoring/metrics)
• ✅ A RAG or LLM-powered application (ideally with evaluation / observability)
• ✅ A data engineering or ETL project (Airflow / Docker / SQL-heavy)
• ✅ Any open-source contribution or reusable template (e.g., FastAPI + RAG boilerplate)

• 📘 Deeply interested in AI governance, AI security, and how organizations adopt ML responsibly
• 📝 Enjoy technical writing, documentation, and sharing implementation details & architecture choices
• 🌌 Outside of work: reading cosmology and theoretical physics, and occasionally playing competitive chess

• 💼 LinkedIn: https://www.linkedin.com/in/arifbadhon/
• 🌐 Portfolio / Blog: https://arifinfo.net/
• 🐙 GitHub: https://github.com/Arif-Badhon
• 📧 Email: [email protected]

If you're working on ML infrastructure, RAG systems, AI governance tooling, or data-heavy backend services, feel free to reach out, open an issue, or start a discussion. Always happy to collaborate on systems that ship to production.