AgentHack submission type
Enterprise Agents
Name
James Jacob
Team name
DocRobotix
Team members
How many agents do you use
Multiple agents
Industry category in which use case would best fit in (Select up to 2 industries)
Universities/Academy
Complexity level
Advanced
Summary (abstract)
This solution orchestrates Robots, Humans, and AI Agents through UiPath Maestro, the Agentic AI Orchestration Framework to automate the full candidate selection journey. It handles application processing, OMR Answer sheet validation, AI-driven handwritten student essay evaluation, and signature verification. The Bot also crops candidate photos and signatures for personalised communication documents. A human-in-the-loop ensures critical accuracy. Final scores determine automated offer or rejection letters, enhanced with AI-generated essay improvement suggestions, making the process smart, scalable, and student-focused.
Detailed problem statement
The candidate selection process in many educational institutions is manual, time-consuming, and prone to errors. Key challenges include:
- Inefficient application processing due to manual data entry and cropping of photos/signatures
- Lack of scalability to handle high volumes during peak periods
- Inability to detect fraud, such as mismatched signatures
- No automated two-way data matching with candidate records
- Tedious and inconsistent evaluation of handwritten essays, often lacking constructive feedback
- Manual communication processes delay the delivery of personalised outcome letters
These issues lead to delays, low accuracy, inconsistent decision-making, and a poor stressful experience for both staff and candidates, highlighting the need for an intelligent, agentic, automated, and scalable solution.
Detailed solution
eCampusAgent: Comprehensive Agentic Automation for End-to-End Candidate Selection Processing
eCampusAgent is a fully automated solution designed to streamline the entire examination processing workflow. Leveraging advanced Agentic AI technology, it scans and extracts data from application forms, performing initial data validation. Forms that fail to meet predefined criteria are automatically flagged for human review to ensure data integrity.
The system further automates image processing by precisely cropping signatures and photographs from the submitted forms, storing these assets securely within a centralised database. Concurrently, eCampusAgent manages the generation and dispatch of test authority letters to candidates, ensuring timely and accurate communication.
Following the examination, eCampusAgent processes the answer sheets by extracting candidate responses and calculating scores based on established answer keys. It performs signature comparisons to detect potential fraud and verifies candidate identity by implementing a two-way data matching mechanism against information extracted from the application forms.
For the assessment of handwritten essays, eCampusAgent integrates an AI Agent that evaluates the content and assigns scores, which are subsequently aggregated with other assessment results. Candidates receive automated, personalised feedback highlighting areas for improvement.
Finally, eCampusAgent determines candidate eligibility by comparing their aggregate scores against defined cut-off thresholds. Based on this evaluation, it autonomously generates and dispatches offer or rejection letters to candidates. This end-to-end automation significantly enhances the examination process’s efficiency and accuracy, reduces manual errors, and accelerates turnaround times.
Demo Video
Expected impact of this automation
End-to-End Candidate Selection Agentic Automation Benefits:
-
Efficiency: Automates application and answer sheet processing, significantly reducing turnaround time. The Essay Evaluator AI Agent uses advanced Agentic AI to assess handwritten essays and provide improvement feedback to each student.
-
Agentic Automation with Maestro: Seamlessly orchestrates AI Agents, robots, and human-in-the-loop collaboration for smarter automation.
-
Accuracy: OCR and automated validation minimise manual data entry errors and ensure precise scoring.
-
Cost Reduction: Cuts operational costs by reducing dependence on manual labour.
-
Scalability: Easily handles high volumes during peak periods without extra manpower.
-
Security: AI-driven signature matching detects and prevents fraud effectively.
-
Better Resource Allocation: Staff can focus on strategic initiatives and student engagement by offloading routine tasks.
-
Consistency: Ensures uniform processing and scoring based on predefined rules.
-
Timely Communication: Automatically generates and sends test authority, offer, and rejection letters to students.
-
Data Management: Centralises extracted data into structured databases for efficient record-keeping and retrieval.
-
Improved Student Experience: Faster, error-free processing and personalised essay feedback enhance overall candidate satisfaction.
UiPath products used (select up to 4 items)
UiPath Agent Builder
UiPath Document Understanding™
UiPath Maestro
UiPath Robots
Integration with external technologies
GMail
Agentic solution architecture (file size up to 4 MB)
Sample inputs and outputs for solution execution
Inputs are stored under ..\Documents folder in each solution
-
Inputs
- Process Indicator = 1 (This is for Application Form Processing)
- Application forms (JPG/PDF/TIFF/PNG)
-
Output
- Cropped Photo & Signature from application forms in a shared folder
- Application outcome letter (with Photo and examination details) into candidate email attachment
-
Inputs
- Process Indicator = 2 (This is for Multiple choice Answer Sheet & handwritten Essay Processing)
- Multiple choice Answer sheets (JPG/PDF/TIFF/PNG)
- Handwritten Essays (JPG/PDF/TIFF/PNG)
-
Output
- Cropped Signature from answer sheets in a shared folder
- Score for correct answers
- Score for essay
- Essay improvement suggestions in the offer / rejection letter
- Signature Comparison score
- 2 way matching result
- Offer letter with Photo, score, & improvement suggestions into candidate email attachment (if selected)
- Rejection letter with Photo, score, & improvement suggestions into candidate email attachment (if rejected)