BiasGuard

Welcome to BiasGuard, a SaaS Platform that applies fairness-aware machine learning algorithms on corporate employee data for organisations to understand biases within their functions and take appropriate corrective actions.

Inspiration

BiasGuard represents a critical response to one of the most pressing challenges in modern organisations - systematic bias. Human biases can influence decisions in hiring, lending, healthcare, criminal justice, and countless other domains. The need for robust bias detection and fairness analysis has never been more urgent.

What it does

BiasGuard is designed to implement bias detection. Its analytical framework and methodologies can be applied to identify, measure, and mitigate human biases within corporate environments. The implemented features are:

• User Registration and Login
• Dataset Upload
• Bias Analysis Engines
• Results Dashboard and Visualisation
• Report Download By leveraging advanced analytics and evidence-based interventions, companies can systematically identify and address the human biases that limit their potential and create more equitable workplaces and decision-making processes.

How we built it

The application was fully built on Bolt.new by vibe coding, seamlessly integrated with Supabase for backend datastore and edge functions, and with Netlify for hosting.

• Frontend - react on Bolt.new
• Backend - Supabase database with Edge functions
• Deployment - Netlify integrated with Bolt.new

Challenges we ran into

The first iteration was attempted with a full set of instructions for Bolt.new to build a complete enterprise platform. The result was a product that looked like a full fledged but with a number of bugs and broken links. Trying to fix them would break unrelated areas which made it difficult to manage it all. A second attempt at breaking up the build in smaller chunks worked better though it ended up in multiple 'repeated automated error' loops. A third fruitful iteration, starting small and building up incrementally, following Bolt.new best practices on effective prompts and minimising token efficiencies resulted in a fully functional application with a good set of features implemented ready to be enhanced to enterprise-grade in incremental steps.

Accomplishments that we're proud of

Standing up a fully functional application with a potential for high impact for organisations to guard against systematic biases, without the need for a single line of developer code!

What we learned

It is possible to dream up bold ideas and build practical products and platforms using the right software, being a little gentle with the tools, and persisting with patience!

What's next for BiasGuard

BiasGuard's potential for systematic bias detection represents a transformative opportunity to create more equitable, inclusive, and effective organizations. The journey toward bias-free organizations requires commitment, transparency, and continuous improvement. BiasGuard provides the tools and framework to make this vision a reality, creating workplaces where every individual can thrive regardless of their background or identity. A set of new features to be implemented are listed:

• Team Management
• Pricing and Subscriptions
• Enhanced Analytics and Reporting
• Compliance and Governance
• Performance and Scalability The next step is to get the product validated by experts and reach out to target organisations to get customer feedback, refine and iterate to achieve Product-Market fit.

The future of work depends on our ability to recognize and overcome our biases. BiasGuard makes that future possible.

Built With

• bolt
• netlify
• react
• supabase