Linear Background

Researching & designing An Applicant tracking system (ATS), designed to find the best software engineering talent

The TLDR

User Problem

Without dedicated recruitment agencies, HR professionals in Sri Lanka struggle to quickly identify qualified software engineering candidates from vast applicant pools.

Team

1 Product Design Lead (Moí)

4 UX Designers

2 front-end engineers

1 Product manager

1 Development team liason

The Process

  • Interviews.

  • Personas

  • User stories.

  • Product feature roadmap.

  • Information architecture map

  • Wireframing design and testing.

  • High-fidelity prototype design.

  • Usability testing.

  • Release v1

The Impact

  • Validated product features, and identified Product-Market Fit (PMF).

  • Implemented Agile RnD processes.

  • Deliver Kastane V1 demo within 8 months.

  • Designed training workshops

Linear Background

Competitive Landscape

There's no end-to-end solution for job applications, hiring, and talent management in Sri Lanka.

  1. Candidates prefer to use job boards like:

    Topjobs.lk , Careers 360 , Ikman , Jobpalm.lk

  2. Companies use in-house solutions that cater to their specific hiring practices.

  3. LinkedIn, Glassdoor, Greenhouse etc. are all unpopular.

Linear Background

User Problem:

Senior staff spend hours/days/weeks helping select applicants even after a recruiter filters candidates

  1. Candidates with impressive resumes frequently fail basic technical assessments.

  2. Technical leadership involvement in resume screening helps catch red flags, but diverts valuable attention from product work.

  3. The high cost of hiring mistakes is painful in the competitive software engineering talent market

Linear Background
Linear Background
Linear Background

Solution

An Applicant Tracking System that verified, and filtered out 5-25% of unsuitable talent

  1. Helps hiring managers quickly bulk filter out under-skilled candidates.

  2. Auto-update candidates on their progress without needing an account.

We had 2 main goals:

  1. Reduce friction for job applicants, while getting the "right" information.

  2. Help hiring managers quickly bulk filter out candidates.


So we designed our final prototype features to cater to our two primary users: job creators, and job applicants. 

Finding the User pain point

The Greatest Pain Point Was Around Bulk Talent Screening, and Screening Out the Most Unqualified Talent

In order to focus our MVP on achieving the strongest “Sean Ellis Test” score, our research goals focussed on identifying the most time/cost-consuming pain-points.


…which was revealed to be the bulk resume screening stage — which many users reported as being time-consuming, costly, and frustrating.

Internal Design Feedback Iterations

We took these new insights and updated the existing design deliverables

We tried to push how far we could improve the quality of applications coming through, while balancing the mental models that recruiters and applicants had collected over time.

  • Linear Background

    Problem

    We opted for a job creation wizard that would break the job creation process into its simplest pieces.

    Feedback

    We overshot simplicity, and discovered a need to stick to the mental models recruiters had gathered over years of sifting through job application tools.

  • Linear Background

    Problem

    Long job descriptions were the standard, but not a good one.


    We explored limiting job criteria creation, encouraging the user to write more targeted job descriptions that attracted the exact talent they were looking for.

    Feedback

    This was a high-friction screen that made no sense to recruiters. We had to find another approach to bridge existing JD's, hiring practices, and a solution that could reliably analyze resumes.

  • Linear Background

    Problem

    Recruiters needed to quickly identify the relevant talent, but have some 'wiggle room' to identify long-shot talent.

    Feedback

    Giving fine control over filtering categories when no applications had come in yet was unrealistic.


    This function would be better served with filters after applications had come through.

  • Linear Background

    Problem

    We had to give them the opportunity to review and correct errors before publishing.

    Feedback

    The high time-on-page for this screen taught us we had too many steps requiring lots of backtracking and editing.

Linear Background

Oversimplification

  • Experimented with a job posting wizard.

  • Over-simplification conflicted with recruiters' established mental models from existing tools.

Linear Background

Balancing novelty with familiarity

  • Initially restricted job criteria creation to encourage more targeted job descriptions.

  • Pivoted away from after recruiters found it cognitively burdensome, and non-transparent.

Linear Background

Grouping Similar functions

  • Initially offered detailed pre-filtering options for candidate selection

  • Shifted to letting recruiters apply filters after applications arrived to reduce cognitive burden.

Linear Background

Error Correction Affordances

  • Added a review step for error correction handling.

  • User-testing showed long time-on-page due to excessive backtracking and editing, indicating need for streamlining.

Key Features

A Job Portal and Applicant Tracking System that Filtered Out 5-25% of the Underqualified Talent Pool

  • Creating & post a Job to all platforms

  • Upload existing Job Descriptions or using a template

  • Parse resumes for exact & fuzzy skills

  • Applicants can apply using their email — no multiple accounts needed

    Linear Background
  • Filter & shortlist candidates in card or row view

    • Linear Background
    • Linear Background
  • Accessibly designed — low contrast, keyboard navigable, with multiple error correction opportunities before and after publishing a job posting

Linear Background

Accomplishments

V1 Launch: Strong Product-Market Validation, 12.5% Usability Gains, and Cross-Team Research Repository

Product Market Fit

Validated & measured core PMF via:

  1. “How would you feel if you could no longer use Kastane” open question.

  2. MOM Test interviews

Usability Issues

  1. Resolved 8-12 usability heuristic violations per screen.

  2. Improved usability test metrics by 12.5%.

Collaborating to ensure Product Delivery

Collaborated with research, design and technical teams to deliver V1 in 8 months.

Improving processes and Talent

  1. Established evergreen research repository with 240 validated entries.

  2. Delivered 6-12 month feature roadmap.

  • Linear Background
  • Linear Background
  • Linear Background
  • Linear Background
  • Linear Background
  • Linear Background
  • Linear Background
  • Linear Background
  • Linear Background

What I Learned

Balancing Business Value, Feasibility, and Team Dynamics

⚖️

Balancing business problems with usability can be tricky and not always overlap.

💻

Including developers in the design process DRASTICALLY reduced design-> dev roadblocks.

Choosing design methods that leverage team strengths (storyboards vs. user stories) is way more important than sticking to the 'right' method.