cost-estimator

# Cost Estimator

Provides frameworks for estimating infrastructure costs, development effort, and total cost of ownership (TCO) for technical projects.

## When to Use

- Planning infrastructure budgets
- Evaluating build vs. buy decisions
- Projecting costs at different scale points
- Comparing technology options by cost
- Creating business cases for technical investments

## Cost Categories

### Total Cost of Ownership (TCO)

```
TCO = Infrastructure + Development + Operations + Opportunity Cost

┌─────────────────────────────────────────────────────────────────┐
│                    TOTAL COST OF OWNERSHIP                       │
├─────────────────────────────────────────────────────────────────┤
│                                                                  │
│  Infrastructure    Development    Operations    Opportunity      │
│  ────────────────  ────────────   ──────────    ────────────     │
│  • Compute         • Engineering  • Support     • What else      │
│  • Storage         • QA           • Monitoring  • could team     │
│  • Network         • DevOps       • On-call     • be building?   │
│  • Third-party     • Management   • Training                     │
│    APIs/SaaS       • Contractors  • Incidents                    │
│                                                                  │
└─────────────────────────────────────────────────────────────────┘
```

---

## Infrastructure Cost Reference

### Cloud Compute Pricing (2024-2025)

#### AWS EC2 On-Demand (US regions)

| Instance | vCPU | RAM | Monthly Cost | Best For |
|----------|------|-----|--------------|----------|
| t3.micro | 2 | 1GB | $8 | Dev/test |
| t3.medium | 2 | 4GB | $30 | Small apps |
| t3.large | 2 | 8GB | $60 | Light production |
| m6i.large | 2 | 8GB | $70 | General production |
| m6i.xlarge | 4 | 16GB | $140 | Medium workloads |
| m6i.2xlarge | 8 | 32GB | $280 | Heavy workloads |
| c6i.2xlarge | 8 | 16GB | $250 | CPU-intensive |
| r6i.2xlarge | 8 | 64GB | $370 | Memory-intensive |

#### GPU Instances

| Instance | GPU | VRAM | Monthly Cost | Best For |
|----------|-----|------|--------------|----------|
| g4dn.xlarge | T4 | 16GB | $380 | Inference |
| g5.xlarge | A10G | 24GB | $730 | ML training/inference |
| p4d.24xlarge | 8x A100 | 320GB | $23,000 | Large model training |

#### Savings Options

| Plan | Savings | Commitment |
|------|---------|------------|
| On-Demand | 0% | None |
| Reserved (1yr) | 30-40% | 1 year |
| Reserved (3yr) | 50-60% | 3 years |
| Spot Instances | 60-90% | Can be interrupted |

### Database Pricing

#### Managed Database (AWS RDS PostgreSQL)

| Instance | vCPU | RAM | Monthly Cost | Connections |
|----------|------|-----|--------------|-------------|
| db.t3.micro | 2 | 1GB | $15 | 50 |
| db.t3.medium | 2 | 4GB | $50 | 100 |
| db.m6g.large | 2 | 8GB | $120 | 200 |
| db.m6g.xlarge | 4 | 16GB | $240 | 400 |
| db.r6g.xlarge | 4 | 32GB | $350 | 500 |
| db.r6g.2xlarge | 8 | 64GB | $700 | 1000 |

**Add for storage**: $0.115/GB/month (gp3)
**Add for IOPS**: $0.02/IOPS/month (over 3000 baseline)

#### Redis/ElastiCache

| Node Type | RAM | Monthly Cost |
|-----------|-----|--------------|
| cache.t3.micro | 0.5GB | $12 |
| cache.t3.medium | 3GB | $50 |
| cache.m6g.large | 6.4GB | $100 |
| cache.r6g.large | 13GB | $175 |

### Storage Pricing

| Service | Cost | Use Case |
|---------|------|----------|
| S3 Standard | $0.023/GB | Frequently accessed |
| S3 Infrequent | $0.0125/GB | Backups, archives |
| S3 Glacier | $0.004/GB | Long-term archive |
| EBS gp3 | $0.08/GB | Block storage |
| EBS io2 | $0.125/GB + IOPS | High performance |

### Network Costs (Often Overlooked!)

| Traffic Type | Cost |
|--------------|------|
| Data IN | Free |
| Data OUT (first 10TB) | $0.09/GB |
| Data OUT (next 40TB) | $0.085/GB |
| Inter-AZ transfer | $0.01/GB each way |
| Inter-region transfer | $0.02/GB |
| CloudFront to internet | $0.085/GB |

---

## Development Cost Estimation

### Engineering Cost Framework

```
Development Cost = (Hours × Hourly Rate) × Complexity Factor × Risk Buffer

Hourly Rate (Fully Loaded):
- Junior Engineer: $75-100/hr
- Mid-level Engineer: $100-150/hr
- Senior Engineer: $150-200/hr
- Staff/Principal: $200-300/hr

Complexity Factors:
- Greenfield, known tech: 1.0x
- Existing codebase, known tech: 1.2x
- New technology for team: 1.5x
- Complex integrations: 1.3x
- Regulatory/compliance: 1.4x

Risk Buffer:
- Well-defined requirements: 1.2x
- Ambiguous requirements: 1.5x
- Experimental/R&D: 2.0x
```

### Story Point to Cost Mapping

| Size | Story Points | Hours | Cost (Mid-level) |
|------|-------------|-------|------------------|
| XS | 1 | 2-4 | $200-400 |
| S | 2 | 4-8 | $400-800 |
| M | 3 | 8-16 | $800-1,600 |
| L | 5 | 16-32 | $1,600-3,200 |
| XL | 8 | 32-64 | $3,200-6,400 |
| XXL | 13+ | 64+ | $6,400+ |

### Team Cost Calculator

```markdown
## Monthly Team Cost

Engineering Team:
- 2 Senior Engineers × $15,000 = $30,000
- 3 Mid-level Engineers × $10,000 = $30,000
- 1 Engineering Manager × $18,000 = $18,000

Overhead (benefits, tools, etc.): 30%
Monthly Burn: ($78,000) × 1.3 = $101,400

Annual Team Cost: ~$1.2M
```

---

## Build vs. Buy Analysis

### Decision Framework

```
Build vs Buy Decision Matrix:

                    LOW Differentiation    HIGH Differentiation
                   ┌────────────────────┬────────────────────┐
    HIGH Volume/   │                    │                    │
    Usage          │     Consider       │       BUILD        │
                   │      Build         │    (competitive    │
                   │   (cost savings)   │     advantage)     │
                   ├────────────────────┼────────────────────┤
    LOW Volume/    │                    │                    │
    Usage          │       BUY          │       BUY          │
                   │   (no question)    │  (then consider    │
                   │                    │   build if scales) │
                   └────────────────────┴────────────────────┘
```

##