Evaluation Task: AI-Powered Project Insight Dashboard
Objective:
The objective of this task is to evaluate the candidate’s ability to design and build a frontend-first web application while
demonstrating working knowledge of Python backend development, DuckDB-based data handling, and AI API integration.
Through this task, the candidate should demonstrate:
Strong frontend development skills, including layout design, responsiveness, and user interaction, Ability to design clean and
usable UI flows for data upload, preview, and insights, Practical experience with Python backend frameworks (Flask or FastAPI),
Correct usage of DuckDB for storing and querying structured data, Ability to integrate an AI API to generate meaningful insights
from data, Understanding of end-to-end application flow, from user input to data processing and output display, The task focuses on
clarity, correctness, and usability, not on complex features or production-scale systems.
Task:
In this task, you are expected to build a small frontend-focused web application that shows how you work across the full stack.
The main goal is to:
- Design a clean and responsive frontend UI Build a basic Python backend
- Use DuckDB to store and query data
- Integrate one AI API to generate simple insights from the data
What to Build:
Frontend (Primary – 50%):
• Responsive dashboard UI • CSV upload section
• Data preview table • AI insight panel
• Clean layout, validation, loading & error states
• Tech: HTML, CSS, JavaScript, React (preferred)
Backend – Python (35%):
• Flask or FastAPI
• APIs for CSV upload and querying
• Store data in DuckDB • Run at least two SQL queries
AI API Integration (15%):
• Use any AI API (OpenAI, Gemini, HuggingFace, etc.)
• Generate 3 insights from dataset summary • Display results on frontend
Evaluation Criteria:
• UI quality and responsiveness
• Backend API design
• DuckDB usage
• AI integration
• Code cleanliness and README
Time Limit:
• Maximum 5 hours
Submission:
• GitHub repository link
• README with setup steps
• Screenshots of UI
• Sample .env file (no secrets)
Note:
This task is strictly for evaluation purposes only.
Incase of any knowledge :
mailUs @ aidevsrs@[Link]
Submit your response @ LINK