Skip to main content

​
Automation & AI Overview

Traditional sustainability tools were built for manual workflows: export data, upload spreadsheets, click through interfaces. But modern carbon accounting demands more—automated pipelines, multi-organization management, and AI-assisted analysis. Dcycle provides three complementary interfaces to match your automation needs:

REST API

Programmatic access for custom integrations and applications

CLI

Command-line interface for scripts, automation, and power users

MCP Server

AI assistant integration for natural language queries

​
The Automation Gap

Most organizations start with manual processes: 
Manual Workflow (typical)
─────────────────────────
1. Export data from ERP/TMS         → 15 min
2. Transform to required format     → 30 min
3. Upload via web interface         → 20 min
4. Verify and fix errors            → 45 min
5. Generate reports                 → 15 min
                                    ─────────
                            Total:  ~2 hours/month
With automation, this becomes: 
Automated Workflow
──────────────────
1. Scheduled script pulls from ERP  → automatic
2. CLI uploads and validates        → automatic
3. Errors flagged for review        → 5 min
4. Reports generated on schedule    → automatic
                                    ─────────
                            Total:  ~5 min/month

​
When to Automate

Signs you need automation:
  • Uploading more than 100 records manually per month
  • Multiple data sources (fleet, facilities, purchases)
  • Regular data refresh cycles (weekly, monthly)
Solution: CLI bulk upload with scheduled scripts
# Monthly automation example
dc logistics upload /data/viajes_$(date +%Y%m).csv --type requests
dc logistics upload /data/consumos_$(date +%Y%m).csv --type recharges
Signs you need automation:
  • Managing 5+ subsidiaries or business units
  • Consolidated reporting requirements
  • Different teams uploading to different organizations
Solution: CLI with organization switching + centralized scripts
# Consolidate data across organizations
for org_id in $ORG_IDS; do
  dc org set $org_id
  dc vehicle list --format json >> all_vehicles.json
done
Signs you need automation:
  • Data validation before upload
  • Audit trail requirements
  • Integration with existing DevOps pipelines
Solution: CLI in GitHub Actions, GitLab CI, or similar
# .github/workflows/upload-emissions.yml
- name: Upload emissions data
  env:
    DCYCLE_API_KEY: ${{ secrets.DCYCLE_API_KEY }}
  run: dc logistics upload data/viajes.csv --type requests
Signs you need automation:
  • Frequent questions from stakeholders
  • Need for quick insights without navigating UI
  • Sustainability team spending time on repetitive queries
Solution: MCP Server with AI assistant
User: "How do our Q3 emissions compare to last year?"

Claude: Your Q3 2024 emissions were 312 tCO2e, down 18%
from Q3 2023 (380 tCO2e). The reduction is primarily
from Scope 2 (-25%) due to your renewable energy switch.

​
Automation Maturity Model

Most organizations progress through these stages:
1

Level 1: Manual

Characteristics:
  • All data entry via web interface
  • Export/import with spreadsheets
  • Ad-hoc reporting
Typical effort: 2-4 hours/week on data management
2

Level 2: Scripted

Characteristics:
  • CLI for bulk uploads
  • Scheduled scripts for regular data
  • JSON exports for analysis
Typical effort: 30 min/week on data managementKey enablers: CLI, --format json, cron jobs
3

Level 3: Integrated

Characteristics:
  • Data flows from source systems automatically
  • CI/CD pipelines validate and upload
  • Alerts on data quality issues
Typical effort: Exception handling onlyKey enablers: API, webhooks, CI/CD integration
4

Level 4: AI-Assisted

Characteristics:
  • Natural language queries for insights
  • AI-generated reports and analysis
  • Proactive recommendations
Typical effort: Strategic decisions onlyKey enablers: MCP Server, AI assistants

​
Key Automation Features

​
Structured Output

All CLI commands support machine-readable output: 
# JSON for programmatic processing
dc vehicle list --format json | jq '.[] | {plate, emissions}'

# CSV for spreadsheet tools
dc facility list --format csv > facilities.csv

​
Non-Interactive Mode

Skip confirmation prompts for automated pipelines: 
# Delete without confirmation
dc vehicle delete $VEHICLE_ID --yes

# Upload without prompts
dc logistics upload data.csv --type requests --yes

​
Environment-Based Authentication

No credentials in scripts: 
export DCYCLE_API_KEY=your_api_key
export DCYCLE_ORG_ID=your_org_id

# Commands use environment automatically
dc vehicle list

​
Composable Operations

Chain commands for complex workflows: 
# Find high-emission vehicles and export details
dc vehicle list --format json | \
  jq '.[] | select(.emissions_tco2e > 50)' | \
  dc vehicle show --format json

​
Choosing the Right Tool

NeedBest ToolWhy
Bulk data uploadCLIDesigned for batch operations
Custom integrationAPIFull programmatic control
Quick queriesMCPNatural language, no coding
Scheduled jobsCLI + cronSimple, reliable automation
Real-time syncAPI + webhooksEvent-driven architecture
Stakeholder questionsMCPAccessible to non-technical users

​
Next Steps

Reporting Pipelines

Set up automated sustainability reporting

Multi-Organization

Manage holdings and subsidiaries

AI-Assisted Analysis

Use AI for carbon insights

CLI Examples

Practical automation scripts