-
-
Landing Page
Inspiration
The project was born out of the daily chaos of the modern workplace—specifically the high-pressure moment when tasks come down from leadership. In these moments, managers often struggle to recall the full skill sets of their entire team, leading to unfair work distribution or tasks being assigned to the wrong people. We wanted to build a "digital brain" that could instantly match a boss's request to the best available talent, removing human bias and cognitive fatigue.
What it does
Smart Task Delegator is an AI-powered dashboard that automates office management:
Intelligent Matching: It analyzes raw task descriptions and cross-references them against a team database.
Dynamic Scoring: It doesn't just look at skills; it calculates a suitability score based on technical expertise, current workload (to prevent burnout), and real-time availability.
Template Support: It provides "Quick-Fill" dropdowns for common office tasks (like Data Analysis or Cloud Migration) to standardize input.
Visual Analytics: It provides visual feedback through workload progress bars and status badges for clear decision-making.
How we built it
Python & Streamlit: Used to build the frontend and the application logic simultaneously.
Data Layer: designed a relational database containing 25 detailed employee profiles, stored in a CSV format.
The Algorithm: implemented a weighted scoring system to rank employees.
The Math: Suitability is determined by the formula:
$$ Z = \sum (S_i \cdot w_s) + A_b - (L \cdot w_l) $$ Where:
\(Z \) is the final suitability score.
\( S_i \) is a binary match for required skills.
\( w_s \) is the Skill Weight (10).
\( A_b \) is the Availability Bonus (up to 5).
\( L \) is the Current Load (0–10).
\( w_l \) is the Load Penalty Factor (0.5).
Challenges we ran into
Weight Balancing: Initially, the algorithm favored "High Availability" too much, suggesting unskilled workers just because they were free. We had to recalibrate the weights to ensure skill match remained the primary driver.
Data Integrity: Standardizing skill tags was essential; We had to ensure the AI wouldn't miss a match due to simple case-sensitivity or spelling variations.
Zero-Base Learning: Learning how to deploy a web-based GUI and connect it to a data backend .
Accomplishments that we're proud of
Burnout Prevention: Successfully building a "Load Penalty" into the code, which ensures that the most skilled worker isn't "punished" with 100% of the office workload.
What we learned
The Logic of Management: Building this taught us that effective management is often a balance of variables—skills, time, and energy—that can be modeled mathematically.
MVP Mindset: We learned the value of the "Minimum Viable Product"—focusing on core utility first .
What's next for Office Task Manager
Direct Integration: Connecting the app to Slack or Microsoft Teams so tasks can be delegated directly from a chat thread. Historical Learning: Implementing a feedback loop where the AI learns from past successes to improve future recommendations.
Log in or sign up for Devpost to join the conversation.