Aur.a

• 

  Aru.a screenshot

Inspiration

The South African job market is competitive and unforgiving. Every application counts, yet thousands of job seekers submit generic, poorly written cover letters simply because they do not know how to present themselves on paper. We built Aur.a because we believe that your writing ability should never be the reason you do not get an interview. Everyone deserves a professional, tailored cover letter that reflects their real potential — regardless of their background or education level.

What it does

Aur.a is an AI powered career tool that helps job seekers generate professional, personalised cover letters in seconds. Users upload their CV, paste a job description, and Aur.a produces a tailored cover letter written specifically for that role. Alongside the cover letter, Aur.a provides an ATS compatibility score that shows how well the CV matches the job description, a CV summary for quick reference, and an AI chat panel where users can refine and improve the generated letter through a live conversation with the AI. Every generated letter is saved to the user's history so they can revisit and reuse past applications at any time.

How we built it

Aur.a was built as a full stack web application using a modern TypeScript stack. The frontend uses a dark professional design system built around the Inter font and a carefully chosen color palette designed to feel premium and trustworthy. Authentication handles user accounts and scopes all data per user. All AI features are powered by OpenRouter using the free model tier accessed through a native TypeScript fetch integration with multi turn reasoning support, enabling the AI to maintain context across the cover letter refinement chat conversation. The application stores user data, generation history, and account details in a structured database. The entire platform was designed and built during the hackathon from the ground up.

Challenges we ran into

The most significant challenge was implementing the multi turn AI conversation correctly. Preserving the reasoning details from each assistant response and passing them back unmodified into subsequent API calls required careful handling to ensure the model could continue its reasoning chain across messages without losing context. Extracting clean usable text from uploaded PDF CVs while keeping the content structured enough for the AI to process accurately was another technical challenge. We also had to balance feature ambition with hackathon time constraints, making deliberate decisions about what to build now and what to plan for later.

Accomplishments that we're proud of

We are proud of delivering a fully functional end to end product within the hackathon timeframe with no mock data and no placeholder features. The cover letter generation works in real time, the ATS scoring produces genuine results, and the multi turn chat refinement maintains live reasoning context across the conversation. We are also proud of the design quality — Aur.a does not look or feel like a hackathon project. It looks like a product that real job seekers would trust and use daily. Most importantly we built something that solves a real problem for a real audience.

What we learned

We learned how to work effectively with OpenRouter's reasoning enabled models and how to structure multi turn conversations that preserve reasoning state between API calls. We deepened our understanding of how ATS systems evaluate CVs against job descriptions and what makes a cover letter genuinely effective for a specific role. We also learned that strong product design decisions made early in a build save significant time later, and that constraining scope ruthlessly is the most important skill in a hackathon environment.

What's next for Aur.a

Aur.a is bigger than a hackathon project. The roadmap includes a full subscription model with Free, Pro, and Career Accelerator plans to sustain development. Upcoming features include a job link importer that automatically extracts job descriptions from LinkedIn and other platforms, full ATS score breakdowns with detailed improvement suggestions, an AI powered CV improvement tool that rewrites and strengthens weak sections, a referral programme that rewards users for growing the community, and a Learning Hub with career guides, interview preparation content, and professional communication resources. The long term vision is to make Aur.a the go to career platform for the South African job seeker — a complete suite of tools that supports every stage of the job application journey.

Built With

• medo
• openrouterapi
• typescript