ML Infrastructure Engineer · Cloud Architect

Ejin Biju.
Cloud infrastructure for
machine learning systems.

Eight years of multi-cloud and Linux engineering, currently focused on MLOps and AI infrastructure. I provision, harden, and automate the platforms that models train and serve on.

  • AWS
  • GCP
  • Azure
  • Hetzner
  • DigitalOcean
  • NovaCloud
§ 01

Background.

Eight years freelancing as a cloud and Linux engineer. VPS provisioning, server hardening, automation scripting, scraping pipelines, bug fixing across stacks I had no business knowing yet learned anyway. Fiverr Level 2 seller, which translates to: clients return, and the work has to ship.

The infrastructure side is the easy part now. What sits next is harder and more interesting: training infrastructure, model serving, feature stores, the boring scaffolding under every AI demo that has to not fall over at 3am. That's the lane.

Studying for AWS SAP-C02. Building MLOps projects in public. Open to architecture roles where the brief is "make the ML platform not collapse."

§ 02

Stack.

Technical stack grouped by category
cloud platforms AWS Google Cloud Azure Hetzner DigitalOcean NovaCloud
systems & devops Linux VPS hardening Bash Docker Git GitHub Actions Automation Web scraping
languages Python Golang Bash
currently building MLflow SageMaker Terraform FastAPI Airflow
§ 03

Projects.

  1. P/01

    GoldenDB

    LIVE

    Offline-first, embedded local database for Go desktop apps. A WatermelonDB alternative engineered specifically for the Fyne framework. Sync layer optional; the local store works alone.

    • Go
    • Fyne
    • SQLite
    • Embedded

    github.com/ejinbiju/goldendb

  2. P/02

    Calorie Tracker

    LIVE

    Personal health tracking application. Logs intake, computes macro targets, persists locally.

    • Python
    • FastAPI
    • SQLite

    github.com/ejinbiju/calorie-tracker

  3. P/03

    ML Training Pipeline

    PLANNED

    End-to-end MLflow pipeline on Terraform-provisioned infra. Experiment tracking, model registry, automated retraining on data drift signals.

    • MLflow
    • Terraform
    • Airflow
    • AWS

    Q3 2026

  4. P/04

    Model Serving API

    PLANNED

    FastAPI inference service with autoscaling on SageMaker endpoints. Request shadowing for canary deploys, latency budget enforcement at the gateway.

    • FastAPI
    • SageMaker
    • Python
    • Docker

    Q4 2026

§ 04

Certifications.

Certifications, in progress and studied
code name vendor state
SAP-C02 Solutions Architect Professional AWS in progress
MLA-C01 Machine Learning Engineer Associate AWS studied
SAA-C03 Solutions Architect Associate AWS studied
AIF-C01 AI Practitioner AWS studied
CLF-C02 Cloud Practitioner AWS studied
§ 05

Contact.

Hiring for ML platform, MLOps, or cloud architecture roles. Open to contract architecture work and one-off infrastructure problems.