GoodSpace is an AI-powered recruitment platform

GoodSpace operates in a highly competitive ecosystem:

Opportunity: Replace resume-heavy workflows with AI-driven decision systems

Through secondary research and product understanding, a few patterns became clear.

Job seekers were not struggling to find jobs, they were struggling to find the right jobs. Nearly half of users reported difficulty identifying relevant roles, and a large portion admitted applying to mismatched positions.

At the same time, recruiters were overwhelmed. A majority reported spending significant time filtering applications, many of which were underqualified.

Another major gap was lack of feedback. Most users never received updates after applying, which reduced trust and increased drop-offs.

Instead of treating hiring as a pipeline, it became clear that it behaves more like a decision system.

To address these issues, we focused on improving signal quality across the journey — even if it meant introducing complexity in certain areas.

One of the biggest decisions was introducing AI-based shortlisting. While this significantly reduced recruiter effort, it also introduced a trust problem. To balance this, we exposed signals like skill match scores and application status, helping users understand how decisions were made.

Another key decision was enabling instant AI interviews. This reduced hiring time drastically, but we noticed users felt pressured. To address this, we added flexibility by allowing users to reschedule interviews, balancing speed with comfort.

We also moved away from traditional keyword matching and introduced skill-based matching. Instead of simply matching resumes to job descriptions, we focused on how well a candidate’s skills aligned with the role.

Across all decisions, one principle remained consistent:

Automation should reduce effort, not reduce clarity

The hiring experience was redesigned into a structured, three-stage system — each stage improving signal quality and reducing friction.

1. Discovery → Finding Relevant Jobs

Instead of overwhelming users with options, we introduced personalized job recommendations with visible skill match indicators. This helped users quickly understand their fit for a role and reduced irrelevant applications.

2. Evaluation → AI Shortlisting

Applications were evaluated instantly using AI, removing the need for manual filtering. Users could now clearly see their application status — whether they were shortlisted, under review, or rejected — reducing uncertainty significantly.

3. Decision → AI Interviews

Shortlisted candidates could proceed directly to AI-powered interviews that took around 10–15 minutes. The experience was designed to feel conversational, with clear instructions and the ability to reschedule if needed.

Rather than treating each feature independently, the platform was designed as a connected system:

This created a reinforcing loop where improving one stage strengthened the entire system.

After rollout, the impact was visible across key metrics.

The overall hiring process became significantly faster, with a 47% reduction in time required from application to interview completion.

User engagement also improved, with an 8% increase in application interactions, indicating that users were more confident in the process.

Most importantly, drop-offs decreased by 12%, showing that reducing uncertainty had a direct impact on user retention.

This project shifted my approach from designing individual flows to designing systems that enable better decisions.

If I were to take this further, I would focus on improving AI explainability and making the system more adaptive to different user behaviors.