Create an AI system that generates long-term workforce sustainability insights by analysing hiring patterns, skills gaps, and potential biases. The tool should help organisations plan a balanced, future-ready workforce using fair, unbiased data models that support sustainable growth.
Organisations increasingly recognise that sustainability is not only about the environment, but also about people. A workforce that is unbalanced, over reliant on a narrow set of skills, or affected by hidden biases can limit long term growth and resilience. However, many companies lack clear, data driven insight into their hiring patterns, skills gaps, and potential biases.
Your task is to create an AI system that generates long term workforce sustainability insights by analysing hiring patterns, skills gaps, and potential biases. The tool should help organisations plan a balanced, future ready workforce using fair, unbiased data models that support sustainable growth.
Prototype or MVP:
A functional web app or dashboard that:
- Ingests workforce related data, such as roles, departments, hiring history, promotions, attrition, and skills
- Identifies skills gaps, concentration risks, and potential patterns that indicate bias in hiring or progression
- Provides clear insights and recommendations to improve workforce balance, diversity, and long term capability
Pitch Deck:
Explain:
- The overall concept and why workforce sustainability is critical for long term organisational health
- How your system turns raw HR or workforce data into meaningful insights for leadership and HR teams
- The AI or analytical models used to detect skills gaps, trends, and potential bias while aiming to minimise unfair or discriminatory outputs
- The potential real world impact, such as better planning, fairer hiring, stronger teams, and more resilient business growth
Technical Summary:
-
System architecture overview:
Workforce data input → preprocessing and anonymisation or aggregation → AI or analytics engine for patterns and risk scoring → insight and recommendation layer → user dashboard
-
Dataset source and preprocessing steps:
Types of HR or workforce data used, handling of missing or noisy records, anonymisation strategies, feature engineering for skills, tenure, departments, and diversity attributes where appropriate and lawful
-
Example of user output:
- "There is a significant skills gap in data and AI related roles over the next 3 years based on current hiring and retirement trends."
- "Promotion patterns in Department X show a possible bias risk. Recommendation: review criteria and introduce structured evaluation to improve fairness."
GitHub Repository:
Include:
- A clear structure for data ingestion scripts, analysis or modelling code, and frontend components
- A README with setup instructions, dependencies, and guidance on how to connect to example or synthetic HR datasets
- A short explanation of model choices, fairness considerations, and how to adapt the tool to different organisations or regulations
Optional:
Add:
- A scenario planning tool that lets organisations test different hiring or training strategies and see projected changes in skills coverage and workforce balance
- A fairness and bias report view that highlights key metrics, explains limitations, and supports transparent communication with stakeholders about how AI is used in workforce decisions
| Category | Description | Weight |
|---|---|---|
| Innovation | Originality of the idea and creativity in approach | 30% |
| Technical Implementation | Quality of the model, data analysis, and functionality | 30% |
| Impact | Practical usefulness, scalability, and potential real-world benefits | 30% |
| Presentation | Clarity, structure, and communication of the solution | 10% |
- Lovable – AI pair programmer that helps you rapidly build and deploy full-stack apps.
- ChatGPT (OpenAI) – Great for brainstorming, debugging, writing explanations, and improving documentation.
- Claude (Anthropic) – Excellent for refining text, structuring ideas, and reviewing project summaries.
- Google Gemini – Helpful for quick code generation, technical research, and idea validation.
- Perplexity AI – AI-powered search tool for real-time information and research support.
- Hugging Face – Access open-source models for NLP, computer vision, and ML.
- GitHub Copilot – AI assistant in VS Code that accelerates coding and debugging.
- Google Colab – Cloud-based Python environment for data analysis and model training.
- Replit – Browser-based coding environment for rapid prototyping and collaboration.
- GitHub – Manage code, track changes, and collaborate with your team easily.
- Kaggle Datasets – Thousands of public datasets across domains, including energy and sustainability.
- UCI Machine Learning Repository – Classic datasets for experimentation and analysis.
- Google Dataset Search – Find open datasets from universities, governments, and organisations.
- Data.gov.uk (Energy) – UK government energy and environmental datasets.