Polo Legacy — Pages, Outputs & User Journeys

Page Map

/                        Home — overview, doc links, env indicator
├── /data                Resource browser (lazy-loaded subsections)
│   ├── /data/users          Cognito users + signup analytics
│   ├── /data/accounts       Customer accounts (team, org, audience)
│   ├── /data/clusters       K8s clusters (dev cell, cloud version)
│   ├── /data/indexes        Marqo indexes (config, price, state)
│   ├── /data/instances      EC2 instances (running only)
│   ├── /data/volumes        EBS volumes (zone, size, state)
│   ├── /data/vpcs           VPCs
│   ├── /data/subnets        Subnets
│   ├── /data/elbs           Load balancers
│   ├── /data/nats           NAT gateways
│   ├── /data/ips            Public IPv4 (alive only)
│   ├── /data/buckets        S3 buckets (size in GB)
│   ├── /data/notebooks      SageMaker notebooks
│   ├── /data/other          Miscellaneous resources
│   ├── /data/savingsplans   AWS Savings Plans
│   ├── /data/prices         Instance pricing reference
│   ├── /data/resources/{code}_{id}       Resource detail page
│   ├── /data/resources/{code}_{id}/tags  Tags table
│   └── /data/resources/{code}_{id}/children  Child resource tree
├── /stats               Analytics (mostly stubs)
│   ├── /stats/clusters
│   ├── /stats/indexes
│   ├── /stats/role
│   ├── /stats/category
│   ├── /stats/customer-compute
│   └── /stats/data-quality
├── /actions             Cleanup recommendations
│   ├── /actions/prune_clusters     Empty/orphaned clusters
│   ├── /actions/prune_volumes      Unattached EBS volumes
│   ├── /actions/prune_ips          Unused public IPs
│   ├── /actions/prune_nats         Unused NAT gateways
│   ├── /actions/stop_notebooks     Idle SageMaker notebooks
│   ├── /actions/dev_indexes        Persistent dev indexes
│   ├── /actions/unroled_instances  EC2 without IAM roles
│   └── /actions/orphaned_instances Instances without parent index
├── /dashboards          Cost reporting
│   ├── /dashboards/cost            Multi-dimensional cost breakdown
│   └── /dashboards/cost_series     Cost trends over time
├── /budget              Budget tracking
│   └── ?where_aws_id=...           Per-account filter
├── /diagram             Interactive infrastructure visualization
└── /query               Freeform SQL-like data queries

Core User Journeys

1. Cost Reduction Workflow

Goal: Identify and eliminate infrastructure waste.

/budget → review spend by account → select account from dropdown
    → compare current 7d avg vs prev 21d avg vs golden day
    → spot increases in usage type breakdown
    → navigate to /actions
        → /actions/prune_clusters → find empty clusters, see savings estimate
        → /actions/prune_volumes → find unattached volumes
        → click resource link → /data/resources/{code}_{id}
            → review owner, cost, timestamps, tags
            → click Jira link to create cleanup ticket

2. Infrastructure Exploration

Goal: Understand what's running and how resources relate.

/data → select resource type (e.g., Instances)
    → sortable table: cost, state, account, role, created_at
    → click instance ID → resource detail page
        → metadata: type, region, path, state, cost, role
        → expand tags section
        → expand child resources (volumes, IPs)
        → view CloudWatch metrics
    → navigate to parent cluster or index via path links

3. Cost Analysis & Reporting

Goal: Understand cost distribution across teams and systems.

/dashboards/cost → income statement card (revenue, gross cost, savings, net)
    → team breakdown cards (DP, CP, S&M, MT, OS, AS)
    → filter by dimension: AWS Account, Type, Audience, Role, System, Team
    → pie charts + detailed tables per dimension
    → sort by cost descending to find top spenders
    → /dashboards/cost_series → line charts of cost trends over time

4. Infrastructure Visualization

Goal: See the full topology and spot anomalies.

/diagram → interactive GoJS canvas
    → toggle resource types: EBS, ELB, NAT, Admin, Index (checkboxes)
    → change detail level: Environments → Accounts → Clusters → Indexes
    → toggle cost overlay (highlights expensive nodes)
    → explore relationships between resources visually

5. Cloud Operations / Alerting

Goal: Automated detection of infrastructure issues (runs on schedule, not user-initiated).

lambda_cop (EventBridge) →
    → find persistent dev indexes (high cost in dev)
    → find instances without IAM roles
    → find orphaned index instances
    → build Slack message blocks
    → post to Slack webhook

Page Detail: Budget

The most complex UI page. Sections:

1. Account selector — dropdown with priority ordering: Total, prod, staging, open_source, controller, preprod, then alphabetical others
2. Spend summary — total 7d spend, daily average
3. Current vs Previous — 7d avg vs prev 7d / prev 21d / golden day, with % change
4. Raw EC2 costs — pre-savings-plan EC2 spend with same comparisons
5. Run-rate vs target — projected monthly, target ($79,167), % to target
6. Daily gaps — current vs golden per day, current vs target per day
7. Usage type breakdown — table sortable by: current avg, prev avg, $ delta, % delta
8. Account breakdown — same structure per account
9. Raw EC2 by account — current vs prev 21d vs golden

Color coding: green = cost decrease (good), red = cost increase (bad).

Page Detail: Cost Dashboard

Structured as an "income statement" with drill-down:

• Income Statement: Total Revenue, Gross Cost, Total Savings, Net Cost, Net Profit
• Team cards: DP, CP, Sales & Marketing, Marqtune, OpenSource, AppliedScience
  • Each shows: Total, Customer portion, Dev portion, week-over-week delta
• Dimension breakdowns (each with pie chart + sortable table):
  • By AWS Account
  • By Resource Type
  • By Audience
  • By Role
  • By System
  • By Team
• Filter via query param where_aws_id, where_type, etc.

Page Detail: Actions

Each action page shows:

• Title and description
• Jira ticket link for context
• Summary: "Delete N resources, save $X.XX/day"
• Sortable table of candidates with: ID (linked), environment, account, owner, cost, updated_at, related resource IDs

Cost Display

All costs stored as $/hr. A header dropdown lets users toggle display between:

• Hourly (raw)
• Daily (x24)
• Monthly (x730)
• Yearly (x8760)

Implemented via CSS variables (--cost-H-display, etc.) toggling visibility of pre-rendered spans.