Hardik Gandhi
Site Reliability Engineer
5+ years exp
Python
Bash
Node JS
JavaScript
C#
Data Pipeline & Feature Engineering: 4/10
Experimentation & Evaluation: 4/10
MLOps & Deployment: 4/10
Active 11 hours ago
Download CVCV
Overview
Technical skills
Projects
Timeline
Roles
Overview
A pragmatic ML engineer (middle level) who builds end-to-end classical ML workflows and utility computer vision tools using OpenCV and scikit-learn. The strongest proven skill is delivering simple MLOps-capable models and evaluation flows, demonstrated by the training script that logs to MLflow and the evaluate.py that computes metrics and saves plots. There is little evidence of advanced production hardening, automated tests, GPU or efficiency engineering, or research-grade model development in public code.
Phone
Technical skills
Languages
5
Python
Bash
Node JS
JavaScript
C#
Node JS
4
Sharp
Axios
Express
Mongoose
DevOps
23
AWS
Kubernetes
Grafana
Docker
GitLab CI
Kibana
Terraform
Ansible
Prometheus
Helm
Jenkins
Logstash
OpenTelemetry
GitHub Actions
CloudFormation
Loki
Azure
Amazon S3
Amazon CloudWatch
Amazon EC2
API Gateway
AWS Lambda
IAM
Databases
4
ElasticSearch
Redis
PostgreSQL
DynamoDB
AI/ML
6
OpenAI
OpenCV
NumPy
Pandas
Scikit-learn
MLFlow
Other
28
SonarQube
Trivy
Pydantic
Marshmallow
.NET
Amazon EKS
GitHub
Git
GitLab
Linux
ArgoCD
CI/CD
Unix
MLOps
CI/CD
DevOps
Frontend
UI
GitOps
TCP/IP
DNS
kubectl
Platform Engineering
Blue-Green Deployment
Configuration Management
Incident Management
SLI/SLO/SLA
Progressive Delivery
Projects
Serverless Data Pipeline
Built an event-driven serverless pipeline where API Gateway triggers a Lambda function to enrich payloads, store raw data in S3, and record metadata in DynamoDB. Added a second S3-triggered Lambda to transform and promote records from raw to processed storage while updating status in DynamoDB automatically. Defined the full AWS architecture as code and automated repeatable deployments via a GitHub Actions CI/CD workflow.
Provisioned a small Kubernetes cluster on AWS EC2 and deployed a multi-service application via Argo CD GitOps sync. Set up Helm-based monitoring using the Kube Prometheus Stack and wrote PromQL queries to track key resource and network metrics across namespaces. Designed the workflow so deployments follow version-controlled manifests with minimal manual intervention and safer rollbacks.
Implemented a multi-stage CI workflow that runs tests, static analysis, builds Docker images, scans for vulnerabilities, and blocks non-compliant deployments. Added GitOps-based CD with Argo CD to reconcile Kubernetes state on image changes, with automated access for a private registry. Included Prometheus and Grafana monitoring to track pipeline throughput and deployment health.
Timeline
Site Reliability Engineer
·
Middle
Stripe
·
Full-Time
Designed AI-assisted incident response workflows using Kubernetes and AWS, integrating logging correlation and runbook/root-cause suggestions to reduce time to resolution. Built highly available multi-region Kubernetes infrastructure on AWS using Terraform, Helm, Argo CD, and CI automation, supporting high-volume payment workloads. Implemented SLI/SLO monitoring and observability dashboards with Prometheus, Grafana, OpenTelemetry, and Loki, and improved deployment safety through progressive delivery and rollback mechanisms. Automated provisioning and access policy controls to speed up environment setup and reduced infrastructure provisioning effort significantly.
Python
OpenAI
Kubernetes
AWS
Terraform
Helm
GitHub Actions
Prometheus
Grafana
OpenTelemetry
Loki
CloudFormation
IAM
Redis
Carleton University
Master's Degree ·
Computer Engineering
Charotar University of Science and Technology (CHARUSAT)
Bachelor's Degree ·
Information Technology
DevOps Engineer
·
Middle
Zoho
·
Full-Time
Built enterprise CI/CD pipelines using Jenkins, GitLab CI, Docker, Kubernetes, and Ansible to automate releases and improve delivery cycle times across production applications. Provisioned secure AWS infrastructure with Terraform and deployment templates, and delivered centralized monitoring and alerting with Prometheus, Grafana, and the ELK stack. Automated server configuration, patching, and operational workflows via Ansible, Python, and Bash to improve consistency. Integrated security and code-quality scanning (SonarQube, Trivy, dependency/secret checks) into pipelines and deployed scalable Kubernetes-based microservices.
Jenkins
GitLab CI
Docker
Kubernetes
Ansible
AWS
Terraform
Prometheus
Grafana
ElasticSearch
Logstash
Kibana
Python
Bash
SonarQube
Trivy
Helm
Middle AI/ML Engineer
Confidence: Medium ML Engineer
A pragmatic ML engineer (middle level) who builds end-to-end classical ML workflows and utility computer vision tools using OpenCV and scikit-learn. The strongest proven skill is delivering simple MLOps-capable models and evaluation flows, demonstrated by the training script that logs to MLflow and the evaluate.py that computes metrics and saves plots. There is little evidence of advanced production hardening, automated tests, GPU or efficiency engineering, or research-grade model development in public code.
Model Architecture & Training
3/10
How well models are designed and trained
Basic classical ML training and model logging using scikit-learn and MLflow; no custom architectures or advanced training controls.
Evidence
mlops-churn-prediction/src/train.py
Data Pipeline & Feature Engineering
4/10
How data is prepared for models
Clear data cleaning and feature engineering for tabular churn data including one-hot encoding and alignment of input columns to the model.
Evidence
mlops-churn-prediction/notebooks/1_data_exploration.py
mlops-churn-prediction/api/utils.py
mlops-churn-prediction/src/data_loader.py
Experimentation & Evaluation
4/10
How results are measured and tested
Basic experimentation and evaluation pipeline with train/test split, logged metrics, saved model, and plotting of confusion and ROC curves.
Evidence
mlops-churn-prediction/src/train.py
mlops-churn-prediction/src/evaluate.py
MLOps & Deployment
4/10
How models are shipped to production
Simple model lifecycle and serving work: MLflow model save/load and a FastAPI endpoint for predictions; suitable for small-scale deployments.
Evidence
mlops-churn-prediction/src/train.py
mlops-churn-prediction/api/utils.py
mlops-churn-prediction/api/main.py
Computational Efficiency
How efficiently computing resources are used
Not evidenced in public code
Research Depth & Innovation
1/10
Depth of research and new ideas
Practical algorithmic work in image deskewing shows applied CV knowledge but no novel research or paper implementations.
Evidence
deskew-image/deskew_service.py
deskew-image/graphics_service.py
Verified artifacts
Expertise
Computer Vision & Image Analysis· Middle
MLOps & Model Lifecycle· Middle
Industries
Telecommunications· Middle
Technologies
MLOps
Python· since 2021 · Middle
PostgreSQL
Redis· since 2025
Terraform· since 2021
Ansible· since 2021
OpenCV
DynamoDB
Helm· since 2021
MLFlow
GitHub Actions· since 2025
Kibana· since 2021
Loki· since 2025
OpenTelemetry· since 2025
CloudFormation· since 2025
Logstash· since 2021
Prometheus· since 2021
Scikit-learn
GitLab CI· since 2021
Azure
CI/CD
GitOps
ArgoCD
Jenkins· since 2021
Pandas
NumPy
Git
AWS· since 2021
Docker· since 2021
Kubernetes· since 2021
ElasticSearch· since 2021
Grafana· since 2021
Pydantic
Marshmallow
Blue-Green Deployment
Platform Engineering
Configuration Management
kubectl
Amazon EKS
AWS Lambda
Amazon EC2
Progressive Delivery
Incident Management
SLI/SLO/SLA
OpenAI· since 2025
GitHub
GitLab
Amazon S3
IAM· since 2025
Amazon CloudWatch
API Gateway
Linux
Unix
TCP/IP
DNS
Recommendations
- Build and productionize small-to-medium classical ML pipelines that include MLflow experiment tracking, model versioning, and a lightweight FastAPI serving layer.
- Develop utility computer vision preprocessing tools and scripts (deskewing, PDF-to-image rendering, resizing) for document OCR and preprocessing pipelines.
- Implement end-to-end pipelines for tabular problems including robust preprocessing, feature contracts, and reproducible evaluation artifacts.
- Harden deployments with automated tests, CI/CD, config-driven MLflow tracking URIs, and clearer error handling in API endpoints.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Junior Backend Developer
Confidence: Medium Generalist
A Junior web developer focused on building React and Next.js sites with simple server-side data fetching. The strongest proven skill is composing Next.js pages with server-side data fetching as implemented in pages/index.js getServerSideProps. The public code does not show backend schema evolution, production-grade security, scalability patterns, or observability and monitoring practices.
API Design
2/10
How well APIs are designed
Minimal API design evidence: a single server-side data fetch is implemented with no versioning, idempotency, retry/backoff, pagination, or consistent error contract.
Evidence
Portfolio/pages/index.js: getServerSideProps fetch to GitHub API and external Google script without retries, backoff, or structured error handling
Data Layer & Database
Working with databases
Not evidenced in public code
Scalability & Performance
Handling load and speed
Not evidenced in public code
System Architecture
2/10
Overall system structure
Small, conventional Next.js page and component composition; some basic config present but no deliberate service decomposition, config/secrets management, or resilience trade-offs.
Evidence
Portfolio/pages/index.js: imports components and containers (Navigation, Greetings, Skills, etc.) to compose the page
Portfolio/next.config.js: images.domains configured for avatars.githubusercontent.com
Security & Auth
1/10
Protecting data and access
Very limited security artifacts: no auth/authz flow, no input validation at boundaries, and external resources are fetched without clear validation or secrets handling.
Evidence
Portfolio/pages/index.js: fetches external APIs in getServerSideProps without validation
Portfolio/portfolio.js: social links and resume URL embedded in code
Reliability & Observability
1/10
Stability and monitoring
Minimal reliability and observability: a simple try/catch with console.log exists but there are no retries, timeouts, structured logging, metrics, or graceful shutdown patterns.
Evidence
Portfolio/pages/index.js: try/catch around server-side fetch that logs the error and returns empty props as fallback
Expertise
Node.js· Junior
Technologies
C#
Node JS· Junior
.NET
Express
Mongoose
Axios
Sharp
Recommendations
- Develop marketing or portfolio websites and small Next.js SSR pages that need lightweight server-side data aggregation.
- Build small full-stack prototypes that connect Next.js frontends to simple Node/Express backends to solidify end-to-end patterns.
- Add robustness to server-side fetches: implement timeouts, retries with exponential backoff and structured error handling in getServerSideProps.
- Improve security and operations maturity by adding input validation, secrets management, and basic observability (structured logs, metrics)
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Intern Frontend Developer
Confidence: Low Generalist
An early-career generalist focused on basic web and DevOps topics with limited production-grade frontend engineering experience. The strongest proven skill is producing a human-authored static HTML demo page, evidenced by a simple index.html demonstrating a working static site. There is no public evidence of component-based frontend engineering, client-side state management, comprehensive testing, or performance/a11y tooling in original code.
UI Component Architecture
1/10
How interface parts are built
Minimal UI component architecture; a single static HTML file with no componentization or reusable UI primitives.
Responsive & Cross-browser
1/10
Works on all screens and browsers
No responsive or cross-browser strategies beyond a viewport meta tag; no layout system, RTL, or feature detection present.
Performance Optimization
1/10
Speed of the interface
No measurable performance work or code-splitting evident; a single static page with no optimization artifacts.
Accessibility & Semantics
2/10
Usable for everyone
Minimal semantic and accessibility signals (html lang attribute present) but no ARIA, focus management, or accessibility testing.
State Management & Data Flow
1/10
Managing data in the app
No state management or async discipline present; only a static page without client-side data flows or forms.
UX & Visual Polish
1/10
Look and feel quality
Basic visual polish limited to simple headings and background color; no UX patterns like skeletons, loading states, or undo actions.
Technologies
Frontend
DevOps
UI
CI/CD
Docker· mentioned only
Jenkins· mentioned only
Node JS· mentioned only
Recommendations
- Develop and deploy small static sites and demos to build frontend fundamentals, starting with semantic HTML, responsive CSS, and accessible patterns.
- Work on simple CI/CD pipeline demos (for example Jenkins pipelines) and documentation to demonstrate DevOps familiarity and automation skills.
- Implement a small end-to-end example (frontend + lightweight API) with tests and basic state handling to show full feature flow and async discipline.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories: