DevOps & Site Reliability Engineer  ·  Cloud Infrastructure  ·  ML Systems

I'm a Computer Science graduate specialising in DevOps, Site Reliability Engineering, and Cloud Infrastructure. My focus is on building observable, reproducible, and automated systems — from containerised ML pipelines to real-time monitoring stacks.

Currently deepening expertise in cloud security, compliance frameworks, and large-scale MLOps.

Site Reliability Engineering Intern — mthree  (March 2025 – September 2025)

• Completed 40+ hours of structured hands-on training in system administration and CI/CD workflows
• Automated routine operational tasks with Bash scripting, reducing manual effort by 40%
• Deployed Prometheus & Grafana monitoring stack; identified and resolved 3+ production anomalies
• Integrated third-party APIs into internal tooling, improving average response times by 20%

AutoScale-ML — Scalable ML Inference Platform  (Apr 2025 – May 2025)
End-to-end inference platform with real-time observability. Containerised Django + MySQL using Docker Compose, cutting environment setup time by 80%. Built a live metrics dashboard with Chart.js tracking RPS, latency, and error rates.

MLOps House Prediction — California Housing Regression  (Aug 2025)
Full MLOps pipeline for predicting median house values. Covers data versioning with DVC, experiment tracking with MLflow, model serving via FastAPI, CI/CD automation, and production monitoring.

Data-Agnostic Agent — AI-Powered Data Analyst  (Sep 2025)
Modular AI analyst with a Streamlit UI and CLI interface. Automatically cleans, validates, and generates insights from arbitrary structured datasets. Includes a ValidationManager layer to detect model hallucinations and ensure output trustworthiness.

ATS Resume Tracker — Gemini AI  (Jul 2024 – Aug 2024)
Resume enhancement tool powered by Google Gemini. Adopted by 10+ users, improving average ATS match scores by 40% and measurably increasing interview conversion rates.

Insurance Fraud Detection & Analysis  (Feb 2024 – Apr 2024)
Benchmarked 6 ML models for insurance fraud classification. Random Forest achieved 92% accuracy, outperforming all other approaches in the comparison suite.

Languages & Scripting
Python · Java · Bash · Shell · HTML · CSS · JavaScript

DevOps & Infrastructure
Docker · Kubernetes · Jenkins · Terraform · Prometheus · Grafana

Cloud
AWS · GCP

ML & MLOps
MLflow · DVC · Apache Airflow · FastAPI · Streamlit

Databases & Tools
MySQL · Git · GitHub · VS Code · PyCharm · Linux