AgentHack submission type
Enterprise Agents
Name
Gagandeep Singh
How many agents do you use
Multiple agents
Industry category in which use case would best fit in (Select up to 2 industries)
Banking and financial services
Complexity level
Intermediate
Summary (abstract)
The Loan Agentic Process is a smart, end-to-end automation flow designed using UiPath Agents, Maestro, Document Understanding, and GenAI, enabling autonomous and collaborative decision-making for loan processing.
Detailed problem statement
Traditional loan processing systems are heavily manual, slow, and error-prone due to:
Unstructured intake of applications via emails and documents (PDFs, Word)
Inconsistent document formats that require human review
Manual interpretation of applicant intent and loan types
Time-consuming validation of financial details and eligibility checks
Limited visibility and delays when involving business users for clarifications
Disconnected systems (email, APIs, humans, bots) causing bottlenecks
The challenge is to streamline and scale loan application processing by:
Reducing human dependency
Automating understanding of documents
Making intelligent decisions
Integrating all stakeholders (bots, agents, humans) in one seamless flow
Detailed solution
Key Components & Flow:
Email Intake & Intent Recognition
Incoming loan applications (PDFs/Docs) received via Gmail
GenAI analyzes email content to detect intent (e.g., Loan Request, Follow-up)
Document Understanding (DU)
Automatically reads and extracts key data from semi/unstructured attachments
Handles multiple formats like PDFs, Word files, etc.
AI Agent Reasoning
Context-aware decision-making engine built using UiPath AI Agents
Validates input data, checks completeness, and routes accordingly
External API Integrations
Calls credit bureau or salary verification APIs for background checks
Fetches real-time financial info for scoring
Human-in-the-Loop (HITL)
If data is incomplete or flagged, an Action Center task is triggered
Business users review/edit missing or inconsistent details via UI Apps
Decision & Communication
Agent calculates loan eligibility & risk score
Generates personalized approval/rejection response
Sends final email with outcome and next steps
Why Agentic?
Combines autonomous AI reasoning with structured workflow
Ensures compliance, accuracy, and business oversight
Enables faster decisions, minimal manual intervention, and scalable operations
Demo Video
Expected impact of this automation
Implementing the Agentic Process for loan applications is expected to deliver significant business and operational benefits:
Efficiency & Speed
Up to 60–70% reduction in loan processing time
Instant extraction of data from emails & documents using DU + GenAI
Faster decisions with AI agents handling initial triage and validation
Accuracy & Compliance
Improved data accuracy through AI-driven validations
Reduced manual entry errors and compliance risks
Enforced standardized decision-making logic across processes
Enhanced User Experience
Business users looped in only when needed via Action Center
Clear, automated communication to applicants
Reduced turnaround times lead to higher customer satisfactio
UiPath products used (select up to 4 items)
UiPath Agent Builder
UiPath Apps
UiPath Autopilot™
UiPath Document Understanding™
UiPath Integration Service
UiPath Maestro
UiPath Robots
UiPath Studio
Integration with external technologies
PDF Monkey
Agentic solution architecture (file size up to 4 MB)
Sample inputs and outputs for solution execution
Input : Email Request, Payslips(Word), Bank statements(pdf)
Output : Mail(Pdf Attachments)
Other resources