AWS Builder Center

Skip the Queue: Deploy AWS Infrastructure with SmartOps Portal

Tired of waiting on infrastructure tickets? Discover how **SmartOps Portal** enables developers to provision AWS resources through a secure self-service experience, powered by automation, governance, and Infrastructure as Code. Deploy faster, reduce bottlenecks, and let your platform team focus on innovation.

Member of Technical Staff. eInfochips - An Arrow Company

1. Introduction

Every growing team hits the same wall with cloud infrastructure: AWS is powerful, but getting things done is slow. A developer needs a VPC — they open a ticket. A product manager wants a staging website deployed — they wait for DevOps. An engineer copies Terraform snippets from a wiki, prays nothing breaks in production, and hopes someone remembered to log the change.
The result is predictable: bottlenecks, shadow IT, inconsistent environments, and audit gaps that show up only when something goes wrong.
SmartOps Portal is a self-service platform built to break that cycle. It gives every approved user — from seasoned engineers to non-technical operators — a single, governed front door to Amazon Web Services. Instead of juggling the AWS Console, a ticket queue, and half a dozen internal tools, teams work in one place:
  • Describe infrastructure in plain language and walk through a safe, reviewable workflow (YAML diff → Terraform plan → PR approval → apply)
  • Create AWS resources directly from a visual dashboard (EC2, S3, DynamoDB, load balancers, auto scaling, certificates)
  • Deploy applications via static website upload, source-code ZIP, or custom Docker image
  • Track every action in a searchable audit log with export to CSV
This blog is a complete walkthrough of SmartOps Portal — what it does, how to deploy to AWS step by step, where to place screenshots, and an honest look at what it does well and where it has limits. Whether you are evaluating the platform for your organization, writing an internal launch announcement, or onboarding your first users, this document is designed to be copy-ready for publication.

Who this is for: Developers, DevOps engineers, platform teams, IT administrators, and business users who need cloud resources without becoming AWS experts.
What you will learn by the end:
  • How to get access and run your first deployment
  • The three ways to put workloads on AWS through the portal
  • Which screenshots to capture for a blog or presentation
  • The platform's strengths, limitations, and recommended use cases



2. Advantages of SmartOps Portal

SmartOps Portal is not just another dashboard pasted on top of AWS. It is designed around how real teams request, review, and ship infrastructure. Here are the main benefits.

For teams and operators

Advantage
What it means in practice
No ticket queue for routine work
Users request VPCs, scale EC2, or deploy apps themselves — within guardrails — instead of waiting days for a manual change.
Plain-language infrastructure requests
The Chat assistant lets users type "create vpc" or "scale to 2 replicas" instead of writing Terraform or clicking through dozens of console screens.
Governed workflow by default
Every chat-driven change passes through YAML review, plan preview, PR approval, and apply — reducing accidental production changes.
Three deployment paths in one UI
Static websites, code ZIPs, and Docker images are all supported without switching tools.
Built-in user guide
Non-technical users get searchable, step-by-step help inside the portal — no external wiki required.
Full audit trail
Log History records chat messages, plan/apply steps, deployments, and logins — with filters and CSV export for compliance.

For platform and security teams

Advantage
What it means in practice
Admin-controlled access
Signup requires approval; roles separate admins from operators; accounts can be rejected or disabled.
Destructive change protection
Plans that delete or destroy resources block the Apply button automatically.
Input validation everywhere
Forms validate app names, domains, AMI IDs, bucket names, ports, and capacity — on both frontend and backend — before anything reaches AWS.
JWT-based sessions
Authenticated SSP workflows; audit logs tied to usernames.
Region switching
Operators can target eu-west-1, us-east-1, and other regions from the Chat toolbar.
Real-time feedback
WebSocket streaming for plan/apply with REST fallback — users see progress live, not a frozen screen.

For developers and DevOps

Advantage
What it means in practice
Live AWS catalog integration
EC2 modals fetch AMIs, instance types, key pairs, launch templates, and subnets from AWS (with sensible fallbacks offline).
Rich service modals
3D architecture diagrams and field guides in every create/deploy dialog reduce guesswork.
Manage Deployments table
Rerun, edit, and destroy app deployments from a single view — no hunting through console tabs.
Single codebase, one port
Flask backend serves portal, dashboard, APIs, and WebSocket on port 5000 — simple to run locally or package for internal hosting.
Extensible middleware
SSP layer (ssp_api.py, ssp_middleware.py) is separate from direct AWS CRUD — easy to swap simulated plans for real Terraform later.

Compared to alternatives

Approach
SmartOps Portal
AWS Console alone
Ticket-only process
Time to first deploy
Minutes (after approval)
Hours (learning curve)
Days (queue wait)
Change review
Built-in YAML + plan + PR
Manual / none
Email or ticket thread
Audit log
Automatic, searchable
CloudTrail (separate)
Scattered
Non-technical users
Supported via Chat + Guide
Poor fit
Depends on engineer
App deployment
Static + code + Docker
Multiple services manually
Custom pipeline per team



3. Disadvantages & Limitations

No platform is perfect. Being upfront about limitations builds trust in a blog and helps teams set realistic expectations before adoption.

Current technical limitations

Limitation
Detail
Mitigation
Simulated Terraform workflow
Chat plan/apply uses middleware output — not live terraform apply against remote state (yet).
Use Dashboard for real boto3 CRUD; treat Chat as a governed preview workflow until Terraform backend is wired.
AWS dashboard routes are unauthenticated
Direct /api/ec2, /api/s3, etc. endpoints do not require JWT today.
Deploy behind VPN or internal network; add auth middleware before internet exposure.
Deploy flows are partially simulated
App deployments save files locally and return mock URLs — they do not always push to real ECR/ECS/CloudFront in every environment.
Confirm with your team which deploy paths are wired to production AWS in your fork.
Single-server SQLite auth store
Users and audit logs live in ssp_data.db on one machine.
For HA production, migrate to PostgreSQL/RDS and add backup strategy.
No multi-cloud support
AWS only (EC2, S3, DynamoDB, ELB, ASG, ACM, ECS patterns).
Not suitable as-is for Azure or GCP teams.
WebSocket uses polling transport
Socket.IO configured without upgrade — works everywhere but not optimal for high-latency networks.
REST SSE fallback already exists for plan/apply.

Operational considerations

Limitation
Detail
Requires AWS credentials on server
The backend needs aws configure (or IAM role) on the host — operators cannot use the Dashboard meaningfully without this setup.
Admin bootstrap not in user guide UI
First admin password appears in server console on first run — must be communicated securely out of band.
AWS costs still apply
Self-service does not mean free — EC2, S3, and other resources incur real charges; the portal does not enforce budget caps.
Learning curve for Chat workflow
Users must understand YAML diff, plan review, and PR approval — simpler than Terraform CLI, but not zero training.
English-only UI
All labels, guide text, and errors are in English.

When SmartOps Portal is not the right fit

  • Large enterprise multi-tenant IDP — If you need full Backstage, Service Catalog, or org-wide service templates with hundreds of microservices, this portal is a starting point, not a replacement.
  • Production Terraform state management — Until real Terraform/OpenTofu integration is added, do not rely on Chat apply as your sole source of truth for production infra state.
  • Public internet SaaS — Default JWT secret, open AWS routes, and local deploy storage are designed for internal/trusted networks, not public multi-customer hosting without hardening.
  • Teams that never want self-service — If your organization mandates that every change goes through a central platform team with no exceptions, the portal's value is reduced.
Blog tip: Acknowledging these limits in your post makes the article credible. Pair each disadvantage with your team's roadmap or workaround where possible.



4. Getting Started — Access & First Login

Before using the portal, every user goes through a controlled onboarding flow. This keeps the platform secure while remaining easy for non-technical users.

Step 1: Open the portal

Navigate to your SmartOps Portal URL. You land on the sign-in page.


Step 2: Request access (first-time users)

Click Request access and fill in
Field
Rules
Username
3–32 characters, letters, numbers, underscore only
Email
Valid work email
Password
Minimum 8 characters
Confirm password
Must match password
Submit the form. You will see a success message — but you cannot sign in yet.


Step 3: Wait for admin approval

An administrator reviews your request under Users in the sidebar. Once approved, return to the login page and sign in with your credentials.

Step 4: Admin one-time server setup (for operators)

On the machine running the portal:
aws configure # Enter AWS Access Key, Secret Key, region
pip install -r requirements.txt
python backend-api.py # Or run start.bat on Windows

The portal connects to AWS via boto3. Without credentials, the dashboard still loads but uses catalog fallbacks.



5. Portal Overview — Navigation & Layout

After login, you enter the main portal shell: a 3D sidebar on the left and your active workspace on the right.

Left menu modules
Menu item
Purpose
Chat
Infrastructure requests in plain language (Provisioner)
Dashboard
View and create AWS resources (EC2, S3, DynamoDB, ALB, ASG, ACM)
Services
Deploy applications (static site, code ZIP, Docker image)
Log History
Audit trail — search, filter, export CSV
Deployments
Manage deployed apps — rerun, edit, destroy
Users (admin only)
Approve or reject signup requests
Guide
Built-in user guide for non-technical users

Top bar indicators

  • Theme — Dark, Light, or Ocean
  • WebSocket / Live — Real-time plan/apply streaming status
  • API status — Shows if backend and AWS are connected



6. Infrastructure Chat — Deploy AWS Resources with Plain Language

The Provisioner (Chat) view is the flagship feature. It turns requests like "create vpc" or "scale EC2 to 2 replicas" into a governed, reviewable deployment workflow.




The 5-step workflow

Step 1 — Select Service, Account & Region

At the top of the Chat view, choose:
  • Service — EC2, VPC, S3, DynamoDB, ECS, Lambda, and more
  • Account — Production, Staging, Dev, Sandbox (with AWS account IDs)
  • Region — eu-west-1, us-east-1, us-west-2, ap-southeast-1 (changes live AWS region)
Use Quick actions chips for common tasks: Scale Up, Create New, Update Config, Add Monitoring, Secure Resource.

Step 2 — Describe your change

Type your request in the chat box or click a quick action. Examples:
  •  create vpc 
  •  Scale EC2 to 2 replicas with monitoring 
  •  Update instance type to t3.medium 
Press Send or  Ctrl+Enter . The assistant responds with a contextual reply and automatically generates a YAML diff


Step 3 — Review YAML diff

Before any plan runs, you must review the Before / After YAML comparison.
  • Green / added lines — new configuration
  • Red / removed lines — deleted settings
  • Destructive badge — appears if the request includes delete/destroy/remove/terminate
Click Confirm Diff & Generate Plan or Reject to cancel.


Step 4 — Review Terraform plan

The portal streams a Terraform plan in real time (WebSocket with REST fallback):
Terraform v1.7.0
# vpc.main will be updated in-place
~ replicas = 1 -> 2
~ instanceType = "t3.micro" -> "t3.medium"
+ monitoring = true
Plan: 0 to add, 1 to change, 0 to destroy.

Destructive plans block Apply automatically — no accidental deletions.



Step 5 — Approve PR and Apply

Every infrastructure change is linked to a Pull Request for review:
  1. Review the PR title (e.g.  [SmartOps Portal] Update vpc in prod )
  1. Click Approve PR until status shows APPROVED (e.g. 2/2 approvals)
  1. Click Apply Changes
Apply is disabled until: YAML confirmed + plan reviewed + PR approved + non-destructive plan.
Every step is logged to Log History (Infrastructure Chat, YAML Diff, Plan Started/Complete, Apply Started/Complete, PR Approved).



7. Resource Dashboard — View & Manage AWS Resources

Open Dashboard from the sidebar to see your live AWS resources inside an embedded view.

The dashboard displays:
  • EC2 Instances — ID, type, state, launch time
  • DynamoDB Tables — name, status, item count, size
  • Application Load Balancers — name, DNS, state
  • S3 Buckets — name, region, creation date
  • Auto Scaling Groups — name, desired/min/max capacity
  • ACM Certificates — domain, status, type
Each section has refresh and create buttons. Resources are fetched directly from AWS via boto3.



8. Service Catalog — All Available AWS & Deploy Services

Open Services to browse the full catalog of what you can create or deploy.

Each card includes:
  • Category badge (COMPUTE, NETWORKING, STORAGE, DATABASE, DEPLOY)
  • Mini 3D architecture diagram preview
  • Deploy time and cost estimate
  • Feature tags (Quick Launch, Auto Scaling, etc.)
  • Create button opening a guided modal


Available services

Category
Services
Compute
EC2 Instance, ECS Fargate, Lambda Function, Auto Scaling Group
Networking
Application Load Balancer, Network Load Balancer, SSL Certificate (ACM)
Storage
S3 Bucket, Static Website Hosting
Database
DynamoDB Table, RDS Database
Deploy
Static Website, Upload Code & Deploy, Custom Docker Image



9. AWS Resource Creation — Forms with Smart Validation

Every create form validates input before sending to AWS — on both frontend and backend.


EC2 Instance

  • Select OS → pick AMI from live AWS catalog (with fallback)
  • Choose from 12+ instance types with category filters
  • Optional instance name and SSH key pair
  • Validates AMI ID format ( ami-xxxxxxxx )



DynamoDB Table

  • Table name regex validation (3–255 chars)
  • Partition key + optional sort key with type selectors (String, Number, Binary)
  • Billing mode cards with visual selection
  • Live schema preview updates as you type


Auto Scaling Group

  • Launch template fetched from AWS
  • Subnet multi-select with format validation
  • Capacity bar: Min ≤ Desired ≤ Max enforced on submit
  • Warning if fewer than 2 subnets selected

Other resources

  • S3 / Static Website — bucket naming rules (lowercase, 3–63 chars, no consecutive dots)
  • ALB/NLB — subnet ID format, minimum 1 subnet (2+ recommended)
  • ACM Certificate — domain FQDN validation, wildcard support
Errors appear as toast messages with plain-language explanations — no cryptic AWS error codes in the UI.



10. Application Deployment — Three Paths to Production

SmartOps Portal supports three deployment models. Pick the one that matches what you have.


Path A — Static Website (S3 + CloudFront)

Best for: HTML/CSS/JS sites, landing pages, documentation sites.
Steps:
  1. Prepare a folder with  index.html  plus assets
  1. Open Services → Deploy Static Website Hosting
  1. Enter app name (lowercase, hyphens) and domain (e.g.  app.example.com )
  1. Select website folder → Submit
  1. Portal uploads to S3, configures hosting, optionally enables CloudFront CDN
Validation: Requires  index.html , max 50MB per file, valid domain format.

11. Manage Deployments — Rerun, Edit, Destroy

All deployments appear in a table view under the Deployments menu.

Action
What it does
↻ Rerun
Re-deploy the same configuration
✏️ Edit & Rerun
Change name, container port, or platform, then redeploy
✕ Destroy
Remove deployment record and delete uploaded files
▸ Expand row
Show deployment ID, platform, port, ECR URI details
Filter by status (Applied, Error, Pending) or search by name/ID. Click + New Deployment to jump to Services.



12. Audit & Compliance — Log History

Every significant action is recorded in Log History.


What gets logged

Action label
When
Infrastructure Chat
User sends a chat message
YAML Diff Generated / Confirmed / Rejected
Diff workflow steps
Terraform Plan Started / Complete
Plan generation
Apply Started / Complete
Infrastructure apply
PR Approved
Pull request approval
Deployment Created / Rerun / Edit / Destroy
App deployment actions
Login / Signup / User Approved
Auth events

Features

  • Search — by user, action, service, or details text
  • Filter — All actions / Infrastructure workflow / Login & users / Deployments
  • Level filter — Info, Warning, Error
  • Export CSV — download full audit trail
  • Auto-refresh — updates when Log History is open during chat workflow
Admins see all logs. Operators see only their own entries.



13. Built-in User Guide — No External Docs Needed

SmartOps Portal includes a searchable, expandable User Guide accessible from the sidebar or the ❓ Help button

  1. What is SmartOps Portal?
  1. Getting Access (First-Time Users)
  1. How to Deploy to AWS (Step-by-Step) ← key section for blog readers
  1. Finding Your Way Around
  1. Using the Chat Assistant
  1. Using the Dashboard
  1. Deploying Your Application (Services)
  1. Managing Deployments
  1. Activity Logs
  1. For Administrators
  1. Simple Glossary (EC2, S3, Docker, YAML, PR, etc.)
  1. FAQ & Help
The guide is written in plain language — no AWS certification required.



14. Admin & Security — User Approvals & Access Control

Administrators have an additional Users menu item for managing access.
📸 Use screenshot: SS-20 — Admin User Approvals Show pending users table with username, email, request date, and Approve / Reject buttons. Include the pending badge count on the sidebar Users icon.

Security model

Feature
Detail
Signup approval
New users start as pending — cannot login until approved
JWT sessions
24-hour tokens, stored in browser localStorage
Role-based access
admin vs operator roles
Account states
pending, approved, rejected, disabled
Audit trail
Every login, signup, approval logged
No default credentials
Admin password set on first DB init (console output or ADMIN_PASSWORD env var)

Production recommendations

  • Set  JWT_SECRET  environment variable
  • Set  ADMIN_PASSWORD  on first deploy
  • Run behind HTTPS reverse proxy
  • Restrict network access to the portal API
  • Configure AWS IAM with least-privilege policies



15. How to Deploy to AWS — Complete Step-by-Step (Blog Summary)

This section is the action checklist readers can follow after reading the blog.

Prerequisites

  • [ ] AWS account with EC2, S3, ECS, ECR, DynamoDB permissions
  • [ ]  aws configure  run on the portal server
  • [ ] SmartOps Portal running ( start.bat  or  python backend-api.py )
  • [ ] User account approved by admin

Option 1 — Infrastructure via Chat

  1. Sign in → open Chat
  1. Select Service (e.g. VPC), Account (e.g. Production), Region
  1. Type request:  create vpc 
  1. Review YAML diff → Confirm
  1. Review Terraform plan
  1. Approve PR until APPROVED
  1. Click Apply Changes
  1. Verify in Log History and Dashboard

Option 2 — Direct resource via Dashboard

  1. Open Dashboard
  1. Find resource type → click Create
  1. Fill validated form → Submit
  1. Confirm success toast and refreshed list

Option 3 — Application via Services

  1. Open Services
  1. Choose deployment type (Static / Code / Docker)
  1. Fill form, upload files → Submit
  1. Check Deployments for status and URL



16. Architecture & Technology Stack

📸 Use animated GIF: SS-GIF —  screenshots/smartops-architecture-flow.gif  Shows 4 animated scenes: (1) 3-layer system architecture, (2) 7-step chat workflow, (3) 3 deploy paths, (4) end-to-end data flow. Regenerate anytime with  python generate_architecture_gif.py .
For technical readers, SmartOps Portal is built as:

Layer
Technology
Frontend
HTML5, CSS3 (3D transforms), Vanilla JavaScript
Portal shell
portal.html, portal-app.js, portal-ssp.js, portal-ws.js
Dashboard
Bootstrap 5, Chart.js, app.js, app-extended.js, app-deploy.js
Backend
Python Flask + Flask-SocketIO
Auth
JWT (PyJWT), SQLite (ssp_data.db)
AWS
boto3 (EC2, S3, DynamoDB, ELBv2, ASG, ACM)
Real-time
WebSocket (polling transport) + SSE REST fallback
Validation
validation.js (frontend) + validators.py (backend)
Deploy storage
Local uploads/ folder + deployments.json

Request flow (Chat → Apply)

User message
→ POST /api/chat/stream (auth required)
→ POST /api/plan/yaml-diff
→ User confirms diff
→ WS generate_plan / POST /api/plan/stream
→ User approves PR
→ WS apply_plan / POST /api/plan/apply-stream
→ Audit logs written at every step




17. Conclusion

SmartOps Portal addresses a problem most cloud teams know too well: AWS is capable enough for everyone, but accessible enough for almost no one without help. By combining a governed chat workflow, a direct resource dashboard, and three application deployment paths in one audited interface, it turns infrastructure from a bottleneck into a self-service capability — without removing the safety rails that production demands.

What makes it worth adopting

  • Speed without chaos — Users move fast through Chat, Dashboard, or Services, but YAML review, plan preview, PR approval, and validation catch mistakes before they reach AWS.
  • One front door — Login, deploy, manage, and audit happen in a single portal instead of scattered consoles and ticket threads.
  • Inclusive by design — Built-in guides, quick actions, and plain-language errors mean product owners and junior engineers can participate — not just senior DevOps staff.
  • Honest scope — It is an internal self-service platform and IDP starter, not a full replacement for enterprise service catalogs or production Terraform pipelines — and that clarity helps teams deploy it in the right place.

Recommended next steps

  1. Run it locally —  start.bat  or  python backend-api.py , then open  URL. 
  1. Configure AWS —  aws configure  on the server so Dashboard and catalog APIs connect
  1. Approve your first users — Walk through signup → admin approval → first chat request
  1. Capture screenshots — Use the SS-01–SS-22 checklist in Appendix A for your blog or launch deck
  1. Share the GIF — Embed  screenshots/smartops-architecture-flow.gif  in architecture sections
  1. Harden for production — Set  JWT_SECRET , restrict network access, add auth to AWS routes, migrate SQLite to RDS

Final word

The goal of SmartOps Portal is not to eliminate platform engineering — it is to free platform engineers from repetitive tickets so they can focus on standards, security, and automation while everyone else gets a safe, documented path to the cloud.
If your team spends more time fielding "can someone create a bucket?" requests than improving your platform, this portal is built for you. Start with Chat and the User Guide, expand to Dashboard and Services, and grow toward full Terraform integration as your maturity increases.
SmartOps Portal — Self-Service AWS Infrastructure & Application Deployment



18. Getting Started Today

# Clone / open the project
cd selfservice

# Install & start (Windows)
start.bat

# Or manually
pip install -r requirements.txt
aws configure
python backend-api.py

# Open browser

Sign up → ask admin to approve → open Chat → type  create vpc  → follow the 5-step workflow.
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article