Enterprisely.ESG Report Builderfor SMEs

Turn utility bills into structured ESG data in minutes, with evidence snippets and review tools. Import utility bills from customer-authorized portals and generate framework-aligned reports (VSME, GRI, ESRS, SASB) with actionable improvement plans.

Evidence Trail

PLN_Invoice_Jan2026.pdf

Uploaded · 42 KB

SOURCE
AI extractionNova 2 Lite
Extracted Fields4 FIELDS
Electricity (kWh)1,247
CurrencyIDR
Cost1,860,450
Billing PeriodJan 2026
Human review

KPI Snapshot

Scope 2 estimate — 986.4 kg CO₂e

Data Quality

88%

Data Gap: Scope 1 Emissions

Missing fuel consumption records

AI Drafted EmailTO SUPPLIER

Hi Logistics Team,

To complete our annual ESG reporting, we urgently need the total fuel consumption (liters of diesel/petrol) for the delivery fleet associated with our account in Q4 last year.

Could you please provide these records by next Friday?

Copy Draft

Top Recommendation

Shift grid load to off-peak hours

High ImpactLow Effort
LowEffort
HighESG Impact

Why: Your latest utility bills show 65% of energy usage happens between 2 PM - 6 PM when rates and grid carbon intensity are highest. Moving heavy equipment usage to morning hours will yield immediate benefits.

85%

VSME Readiness

Almost ready to export

Company Profile100%
Evidence Documents12
Open Data Gaps2 pending
nova_act_agent running...Navigating to PLN Portal ✓Authenticating ✓

Automated Import

Automates supported portal workflows with Amazon Nova Act, then retrieves bills, uploads documents, and runs ESG extraction automatically.

Sync complete
Built on
Amazon Nova 2 LiteAmazon Nova Act
Amazon BedrockAWS TextractAWS S3DynamoDB
Next.js 16React 19Tailwind CSS v4

Amazon Nova Integration

Powered by two Nova models.

Enterprisely uses Amazon Nova at its core — Nova 2 Lite for AI document intelligence and Nova Act for autonomous browser automation.

Nova 2 Lite · Document Intelligence

Multimodal ESG Data Extraction

// Multimodal document extraction

model: "eu.amazon.nova-2-lite-v1:0"

input: [PDF, image, invoice]

output: { vendor, period, kWh, cost }

confidence: 97% // shown in review

Every uploaded bill, invoice, or receipt is processed natively as raw images or text by Amazon Nova 2 Lite via AWS Bedrock. The model extracts structured ESG fields with confidence scores, then surfaces them in a review workflow where teams can confirm or correct key values.

  • Parses PDFs, images, and scanned invoices
  • Extracts energy (kWh), cost, vendor, billing period, and currency
  • Confidence score per field, surfaced in review
  • Native multimodal image processing + Textract fallback

Nova Act · Browser Automation

Autonomous Utility Portal Sync

// Browser automation agent

nova_act_agent running...

→ Navigating PLN portal ✓

→ Authenticating ✓

→ Bill downloaded → S3 ✓

Amazon Nova Act powers Portal Connect — an AI browser agent that automates supported portal workflows, retrieves bills, uploads documents, and runs ESG extraction automatically.

  • Navigates real utility portals autonomously
  • Authenticates, locates and downloads the latest bill
  • Uploads PDF to S3 and triggers extraction pipeline
  • Can run monthly or quarterly when Scheduled Import is enabled

How It Works

Seven steps from scattered documents to a shareable ESG report.

01 Upload or Connect

Upload bills, invoices, and receipts, sync from cloud storage, or import automatically via Portal Connect.

PDFs, Images, Others
Running Agent
02 Automate

You can schedule recurring portal imports monthly or quarterly, and enable monthly auto-sync for Google Drive.

03 Extract

Automatically extract dates, costs, usage, and currency from images, PDFs and other documents.

Structured Data
450 kWh / $120.50
Document Trace
Verified Source
04 REVIEW SOURCES

Open the review workflow to inspect evidence snippets and source documents for extracted fields.

05 RESOLVE GAPS

Our model analyzes data gaps, prioritizes the most important actions, and drafts supplier outreach emails to resolve them.

Missing Data
Draft: Supplier Request
Target Met
06 MONITOR METRICS

Track KPI rollups and estimated Scope 2 emissions as data flows in.

07 Export & Save

Export framework-aligned reports (VSME, GRI, ESRS, SASB) as professional PDF documents.

PDF Export Ready

What makes it different.
Extract. Review. Report.

Stop manually compiling data. Drop in any bill or receipt, let our AI extract the core metrics, and build a framework-aligned ESG reportfor your business.

AI Document Extraction

Drop in any bill, invoice, or receipt. Our multimodal AI extracts energy, cost, vendor, date, and currency details while reducing manual data entry.

"vendor": "ACME"
"kWh": 1200
"cost": $450

Autonomous Utility Sync

Put data collection on autopilot. Save your utility portals for recurring monthly or quarterly imports, or trigger them on demand.

LOGIN
LATEST BILL

Traceable Audit Trail

Auditable by design. Key extracted fields are paired with evidence snippets and source-document access to support review and auditability.

Total Cost
$1,450.00

Action Plans & Missing Data Detection

Never miss a metric. Our system detects missing ESG data, generates step-by-step improvement plans, and drafts outreach emails to speed up follow-up.

DRAFT READY
SEND

ESG Report Export

Auto-generate framework-aligned ESG reports (VSME, GRI, ESRS, SASB) with environmental, social, and governance sections, plus evidence register and readiness context.

ENV
SOC
GOV

Built to be defensible.

Every extracted data point is traceable, scored, and reviewable,with evidence and data-quality context carried into the reporting workflow.

01

Reporting Scope

Define your applicable frameworks (VSME, GRI, ESRS, SASB) and organisational boundaries before any data is collected.

02

Data Capture

Upload bills or connect utility portals. Structured fields are extracted automatically from PDFs, images, and invoices.

03

Evidence Mapping

Key extracted fields are paired with evidence snippets and source-document access in the review workflow.

04

Confidence Scoring

Each extracted field includes a confidence score, and lower-confidence values are surfaced in the review workflow.

05

Human Review

A dedicated review queue lets you confirm, edit, or recheck AI-extracted values before finalizing reports.

06

Execution Output

Export framework-aligned PDFs with evidence register and data-quality context for internal review or external sharing.