Green Invoice Intelligence Agent - Agent Hack

AgentHack submission type

Enterprise Agents

Name

Naveen Kumar Sridhar

Team name

BotBrains

Team members

Naveen Kumar Sridhar, Arun Rajagopal A, Logeswarran S

How many agents do you use

One agent

Industry category in which use case would best fit in (Select up to 2 industries)

Finance

Complexity level

Intermediate

Summary (abstract)

The Green Invoice Intelligence Agent is an enterprise-grade, AI-powered automation that transforms traditional invoice processing into a sustainability-driven workflow. Triggered by invoice emails, it uses Document Understanding and LLMs to extract data, analyze the carbon footprint of purchased items, and intelligently suggest green-rich alternatives for non-sustainable products.

With integrated human-in-the-loop validation and seamless communication to finance teams, this agent helps enterprises automate ESG compliance, reduce CO₂ emissions, and make smarter, greener procurement decisions — all without disrupting existing processes.

Detailed problem statement

Enterprises process thousands of invoices every month without evaluating the environmental impact of the purchased items. Manual invoice handling lacks visibility into carbon emissions, making it difficult for finance and procurement teams to align spending with sustainability goals. There’s no scalable way to identify non-green items or recommend greener alternatives during invoice review.

This automation addresses the growing need for sustainable procurement by introducing an AI-powered agent that analyzes invoice content, estimates CO₂ emissions, and recommends eco-friendly item replacements. It eliminates manual classification, reduces turnaround time, and empowers organizations to make greener purchasing decisions — automatically.

Detailed solution

Trigger: Starts automatically when an invoice email is received.

Email Handling Bot: Fetches the invoice attachment and reads metadata.

Data Extraction: Uses Document Understanding + LLM to extract vendor, items, and pricing.

CO₂ Analyzer Agent: Calculates carbon footprint and classifies items as Green or Non-Green.

Recommendation Agent: Suggests eco-friendly replacements for non-green items.

Decision Gateway: Routes invoices with green alternatives to human validation (HITL).

Human-in-the-loop: Finance team reviews and approves/edits suggested replacements.

Email Output: Sends a final summary with green alternatives to finance for action.

Audit & Reporting: Logs all data (original, suggested, CO₂ impact) into Excel/dashboard.

Outcome: Automates sustainability compliance, reduces carbon-heavy purchases, and supports ESG goals.

Demo Video

Expected impact of this automation

:white_check_mark: 80% time saved in invoice review and classification

:white_check_mark: 90% reduction in manual/repetitive tasks related to sustainability checks

:white_check_mark: Improved ESG compliance with automated CO₂ emissions tracking

:white_check_mark: Real-time insights into carbon footprint of procurement

:white_check_mark: Higher ROI through smarter, sustainable purchasing decisions

:white_check_mark: Audit-ready logs and reports for traceability and compliance reviews

:white_check_mark: Supports organizational sustainability goals across departments

:white_check_mark: Minimizes errors by eliminating manual data handling and green classification

UiPath products used (select up to 4 items)

UiPath Document Understanding™
UiPath Orchestrator
UiPath Studio

Automation Applications

Gmail

Integration with external technologies

Gmail,Open AI

Agentic solution architecture (file size up to 4 MB)

Sample inputs and outputs for solution execution

1.invoice from mail
2.extracted item list as Agent Input
{ “Recycled A4 paper reams”, “Biodegradable packaging materials”, “HP LaserJet printers”, “CFL bulbs”, “Air freight express delivery charge” } }
3.Alternate Item which is co2 emission free as Agent Output
{ ExpectedCO2Reduction=“30%”, OriginalItem=“Recycled A4 paper reams”, SuggestedAlternative=“Tree-free paper made from agricultural waste” }, content { ExpectedCO2Reduction=“50%”, OriginalItem=“Biodegradable packaging materials”, SuggestedAlternative=“Reusable and returnable packaging systems” }, content { ExpectedCO2Reduction=“25%”, OriginalItem=“HP LaserJet printers”, SuggestedAlternative=“Refurbished or remanufactured printers” }, content { ExpectedCO2Reduction=“70%”, OriginalItem=“CFL bulbs”, SuggestedAlternative=“LED light bulbs” }, content { ExpectedCO2Reduction=“90%”, OriginalItem=“Air freight express delivery charge”, SuggestedAlternative=“Ground shipping or local sourcing” } } }

Other resources