Shaik Afzal Hussain · ML Support Engineer

About Me

Building reliable, scalable cloud systems and shaping responsible AI.

I am an ML Support Engineer at Aptly Technology Corporation, where I provide ML operational support, implement MLOps practices, and help keep models production-ready in production environments. I focus on model health monitoring, incident response, and reliable model deployments.

Previously at Amazon Web Services (AWS), I worked on AWS Bedrock to train, evaluate, and improve Large Language Models (LLMs), operating on multi‑modal datasets (text, speech, audio, image, and video) and improving data quality and evaluation workflows.

Previously I worked as an AWS Cloud Application Developer designing and deploying secure, multi‑AZ, auto‑scaling architectures using EC2, ELB, IAM, EBS, S3 and Auto Scaling, and automating cloud operations with Bash and Python. I also design CI/CD and DevOps workflows (Docker, Kubernetes, Terraform, Jenkins, SonarQube, DockerHub) that deliver reliable releases for cloud and AI workloads.

Selected outcomes include delivering 99.99% uptime and a 45% reduction in latency across production services. I’ve built high‑availability AWS apps, scalable 3‑tier architectures, and end‑to‑end DevOps CI/CD pipelines, and contributed to projects such as CoCreate.AI, Study Buddy AI, and the ATS Resume Scanner.

Hands-on with Amazon Q and Amazon Nova foundation models.
Built secure, multi‑AZ, auto‑scaling cloud architectures on AWS.
Results‑driven: improved uptime to 99.99% and reduced latency by 45%.
Google Arcade Cohort 2 — completed hands‑on labs, quests, and cloud skill challenges on GCP.

Professional Experience

Total experience: —

ML Support Engineer

Aptly Technology Corporation · Client: Microsoft · 03/2026 – Present · Bengaluru, India —

Provide ML operational support, monitor model health, and troubleshoot production issues for Microsoft as the primary client.
Implement MLOps pipelines and assist with model deployment, versioning, and rollback strategies.
Collaborate with data engineers and SREs to ensure reliable, scalable inference services.
Document runbooks, incident reports, and support guides for internal stakeholders and Microsoft engineering teams.

ML Data Associate, AWS Bedrock

Amazon Web Services (AWS) · 11/2024 – 02/2026 · Bengaluru, India —

Train and evaluate LLMs via AWS Bedrock; exposure to Amazon Q and Amazon Nova.
Deliver high-quality labeled datasets across text, speech, audio, image, and video.
Analyze error patterns, propose fixes, and improve data quality and processes.
Operate in-house tooling; interpret and implement detailed technical instructions.

AWS Cloud Application Developer

Rooman Technologies Pvt Ltd · 02/2024 – 10/2024 · Kurnool, India —

Designed and deployed infrastructure with EC2, Auto Scaling, ELB, IAM, and EBS.
Automated ops and admin tasks using scripts, improving efficiency.
Worked across Linux, shell scripting, databases, and data management.

Technical Skills

Cloud

AWS: EC2, ELB, EBS, S3, IAM, Auto Scaling, CloudWatch, SNS, SQS, RDS

GCP: Compute Engine, Cloud Storage, Cloud SQL, Cloud Monitoring, Pub/Sub, IAM, Cloud DNS, Cloud Functions, GKE

DevOps

Git, GitHub, Docker, Kubernetes, Jenkins , Terraform

Scripting

Bash, Python (basic)

Artificial Intelligence (AI)

Language Models (LLM) , Agents , Natural Language Processing (NLP) , Automation , AWS Bedrock Azure AI, Google Vertex AI , Prompt Engineering, Retrieval-Augmented Generation (RAG), Model Evaluation, Fine-tuning, Vector Databases & Embeddings , LangChain

Selected Projects

ATS Resume Scanner

Free GCP-hosted tool to analyze resume ATS-fit, keyword gaps, and feedback.

Docker + Artifact Registry + Cloud Run
~30s analysis · ~$0.50–$1.00/month via free tiers

Live App

High Availability Web App on AWS

Highly available deployment leveraging Auto Scaling, ELB, Multi-AZ.

99.99% uptime · 45% faster loads · +30% responsiveness
+50% concurrent users during peaks

Source

3‑Tier Cloud App on AWS

Scalable 3‑tier architecture in AWS with Auto Scaling and secure IAM.

+40% capacity scaling without manual intervention
99.9% uptime via Multi‑AZ + Auto Scaling

Source

End-to-End DevOps Project — Flask App CI/CD Pipeline

A complete end-to-end DevOps lifecycle project demonstrating CI/CD for a Flask application using Terraform, Jenkins, SonarQube, Docker, DockerHub, and AWS EC2.

Automated build and test pipeline in Jenkins
Static analysis with SonarQube
Docker image build & push to DockerHub
Automatic deployment to AWS EC2
Infrastructure provisioning using Terraform

Architecture Flow: GitHub → Jenkins → SonarQube → DockerHub → AWS EC2

GitHub

CoCreate.AI — Human + AI Co-Creation Platform

An AI-powered web platform built for Vibeathon 2025 that improves human-AI collaboration using prompt refinement, story mode, idea mode, tutor mode, and more.