AgentHack submission type
Enterprise Agents
Track Type
Track 2: UiPath Maestro BPMN
Name
Sagar Patel
Team name
1RTeam
Team members
@zoyebmansuri , @Oozair_Khan , @SagarPatel , @Jeni_Patel
How many agents do you use
Multiple agents
Industry category in which use case would best fit in
Finance, Accounting, Government
Complexity level
Advanced
Summary
We present Tax IQ, a GenAI-powered agentic automation solution built on UiPath Maestro to transform tax notice management. The system ingests notices from email and automatically creates cases, extracts key data, and classifies notice types using AI agents. Based on risk assessment and classification, it intelligently routes cases into appropriate resolution paths such as payment, dispute handling, or compliance actions. A human-in-the-loop validation step ensures accuracy and governance before execution. RPA bots then automate downstream actions across enterprise systems, including notifications and record updates. The solution significantly reduces manual effort, improves accuracy, and accelerates response timelines.
Detailed problem statement
Tax notice management in enterprises is highly manual, fragmented, and time-sensitive, leading to inefficiencies and compliance risks.
- Tax notices are received across multiple channels (emails, portals, uploads) with no standardized intake, making tracking and prioritization difficult.
- Extracting critical information from unstructured notices such as PDFs and emails is time-consuming and prone to human error.
- Classification of notices (payment, dispute, compliance, etc.) depends heavily on manual judgment, causing inconsistencies and delays.
- Risk assessment and decision-making lack structured intelligence, leading to missed deadlines, penalties, or incorrect actions.
- Resolution workflows require coordination across multiple systems (ERP, accounting, portals), creating operational delays.
- Human validation processes are inefficient, often lacking clear visibility into case status and decision rationale.
- Teams spend significant time on repetitive tasks instead of focusing on complex analysis and strategic decision-making.
Detailed Solution
We propose an intelligent, agentic automation solution leveraging UiPath Maestro, agents, and RPA bots to streamline tax notice management end-to-end:
Agentic Automation Pipeline
1. Intake & Case Creation
- Email trigger starts the process
2. Record creation (Data Fabric)
- Creates structured case record in data fabric for tracking and processing
- Uploads files against the record for AI processing
3. Extraction Agent
- Extracts structured details from notices:
- Taxpayer info
- Notice type indicators
- Amount, due date, references
4. Classification & Assessment Agent
- Classifies notice into Payment / Dispute / Compliance / Refund
- Performs risk assessment (high / medium / low)
5. Decision Engine
- Applies branching logic based on classification + risk
- Routes case to appropriate flow
- Flags high-risk cases for escalation
6. Resolution Agent
- Recommends action: Payment / Dispute / No action / Correction
- Prepares decision summary for review
7. Human-in-the-Loop Validation
- User validates agent decision (Approve/Reject) in UiPath action apps
- Ensures compliance and accuracy
8. Execution task
- Notifications (approval/rejection emails)
- Update case records
Expected impact of this automation
The implementation of this agentic automation solution is expected to have the following impacts:
Increased efficiency: Reduces manual processing effort by up to 60–70%, enabling faster handling of tax notices with minimal human intervention.
Improved accuracy and compliance: AI-driven extraction and standardized classification significantly reduce errors and ensure consistent regulatory adherence.
Faster turnaround time: Automated decisioning and routing accelerate response timelines, helping avoid penalties and missed deadlines.
Enhanced visibility: Real-time dashboards of maestro and data fabric provide end-to-end tracking of case status, risk levels.
Scalability: Easily handles high volumes of tax notices across multiple channels without requiring additional workforce.
Proactive risk management: Built-in risk assessment identifies and escalates high-risk cases early for prioritized handling.
Reduced decision bias: By leveraging structured data, predefined rules, and AI-driven classification, the system ensures consistent and objective handling of tax notices, minimizing dependency on subjective human judgment.
UiPath products used
UiPath Agents
UiPath maestro
UiPath Action Apps
UiPath Data Fabric
UiPath cloud studio (RPA Flow)
Integration with external technologies
Outlook
Agentic solution architecture
Link to code: GitHub - Sagar-Patel126/Tax_Notice_Agentic_Process_UiPathAgentHack-2026: Agentic process of tax notice · GitHub