Recruit Assist

Your go to conversational friendly recruitment platform along with AI

UX case study

FIGMA PROTOTYPE

OVERVIEW

Recruiters been spending 29% of their time on screening and administrative work. Creating unclear job descriptions also lead to 62% of employers attracting unqualified candidates and then 45% receiving too few good applicants.

#Solution

An AI-powered recruitment platform automating 3 key workflows: JD generation, skill-based resume screening, and interview/evaluation process.

%Impact

lower screening effort and time-to-fill, better shortlist quality, and fewer unqualified applicants—with validation through 3 enterprise customer demos and a 5-customer pilot goal.
18% → 10-12% screening | 11% → 6-7% admin | 62% → 40-45% unqualified applicants | 3 → 5 customers | 3 AI workflows.

DOMAIN

B2B | Saas

ROLE

Product Designer
(Solo)

DURATION

7 months

TEAM

Design Manager
Product Managers
Developers

My responsibility

Led end-to-end 0→1 design across JD Generation, Resume Screening, Interview Scheduling, AI Feedback, and Analytics.

Explored 3 product concepts (Web App, MS Teams Bot, Browser Extension) before defining the final product direction.
Conducted competitive analysis with PM's across leading recruitment platforms to identify market opportunities.
Presented designs through multiple stakeholder reviews, iterating based on leadership and client feedback.
Showcased interactive prototypes to enterprise clients, supporting product validation and adoption.

The MVP specifically focused on Recruiter + Hiring Manager, based on customer interactions.

WHAT I DID BEFORE STARTING THE FINAL DESIGN

PROBLEM SPACE

Hello world

Working across different platforms

29% Recruiters move across creating unnecessary context switching and administrative effort. How do we use Generative + Conversational AI to reduce repetitive recruiter work while keeping recruiters and hiring managers in control of hiring decisions?

Hello world

Seamless Experience as per persona

Recruitment involves multiple people with very different needs recruiters, hiring managers, candidates and panelists.

Hello world

Not to break the ATS flow

Recruiters already worked inside an ATS — how could we add value without creating another disconnected tool?

Hello world

Turst issues with AI

A recruiter can't simply trust an AI-generated candidate score. They need relevant skills, candidate context and the ability to make the final call.

How might we use AI and reduce all those repetitive recruitment work while keeping hiring decisions human, trustworthy, and seamlessly connected to existing ATS workflows?

SOLUTION

Automating recruitment process

Designed AI-assisted JD generation, candidate screening, ranking and evaluation rather than making AI the final decision-maker.

Designing one experience for different hiring stakeholders

Mapped the end-to-end recruitment journey and created persona-specific experiences, with the MVP focused on Recruiters + Hiring Managers.

AI working with the existing ATS

Explored different integration models and designed easy navigation between RecruitAssist and the ATS, minimizing unnecessary switching.

Making AI recommendations trustworthy and useful

Designed skill-based candidate matching, candidate highlights, skill insights and ranking/evaluation workflows, with a focus on reducing reliance on traditional keyword matching and addressing bias.

IMPACT

Screening effort

18% → 10-12%

Time reduction

Administrative effort

11% → 6–7%

Time reduction

Unqualified applicants

62% → 40–45%

Filtered easily

Demos

3 → 5 (now 8 clients)

Enterprise demos → pilot target

MVP

3 core AI workflows

Recruiter + Hiring Manager

DESIGN PROCESS

Define

Conceptulize

Design

Iterate

Mapped the end to end recruitment journey

User Personas

Lana Ray, 37 - Hiring Manager, Product Team

Lana is a busy hiring manager who collaborates with the recruiting team through all hiring stages to find the perfect candidates for her huge PM department.

Pain points

Receives candidate names but no context no explanation of why someone was shortlisted, no comparative view
Has to write interview feedback from scratch after every session, often hours or days later
No structured way to communicate her decision back to the recruiter it all happens over email or calls.

Goals

She aims to enhance business outcomes aligning with her department's hiring needs and collaborating seamlessly with the recruiting team to achieve swift time-to-hire.

John Scott, 33 - Head of Recruitment

John is a recruiter who regularly collaborates with other stakeholders and uses an ATS to conduct his daily recruitment activities.

Pain points

Writes JDs from scratch for every new role, often starting from a blank page or copy paste from different platforms
Reviews resumes one by one against a mental model of what "good" looks like and sometimes cant see all of them.
Spends hours chasing panelists for calendar availability and sending back-and-forth scheduling emails
Has no real-time visibility into where candidates are in the pipeline unless he manually checks the ATS

Goals

John wishes to best utilize his time by working on high-value activities such as candidate interactions to achieve efficiency, productivity, and engagement for his business.

Competitive analysis - Recruitment

Exisiting Capability

Roadmap

No Capability

Limited Capability

Focused on these Goals

User Goal: Find and hire the right candidate faster.

Business Goal: Reduce time-to-hire and increase recruiter productivity.

User Goal: Automate repetitive hiring tasks with AI.

Business Goal: Lower operational costs and improve recruitment efficiency.

User Goal: Make confident, data-driven hiring decisions.

Business Goal: Improve the quality of hires and increase hiring success rates.

Key trade off - concepts

CONCEPT 1 — Microsoft Teams Bot ⚡ Secondary Channel ❌ Rejected

Recruiters and hiring managers to interact with RecruitAssist within Teams.

Strength: Convenient for hiring managers

Risk: High back-and-forth navigation

CONCEPT 2 — ATS Browser Plugin ❌ Rejected

An AI assistant layered on top of existing ATS platforms like SAP SuccessFactors.

Strength: Less switching apps

Risk: Higher development/testing complexity

CONCEPT 3 — Standalone Web App ✅ Selected

A dedicated recruiter workspace with an AI assistant embedded across the product. Candidate data, jobs, interviews, and actions live in one place, eliminating context switching and creating a seamless hiring workflow.

Strength: Stronger overall expeience, one source of truth for all, required less time to build.

Risk: Some switching between ATS + RecruitAssist

Hypothesis - Concept Comparison for Recruiter Persona

The vision was much broader than the MVP. Rather than trying to solve the entire recruitment lifecycle at once, we narrowed the first release around Recruiters and Hiring Managers and focused on the workflows where AI could create the most immediate value with JD creation, candidate screening and hiring decisions.

USE CASES

Content Generation

Job Descriptions

Candidate Emails
Interview Questions

Smart AI Search & Discovery

Smart Search
Candidate Matching

Screening & Summarization

Resume Parsing
Candidate Summaries
Resume Screening

Ranking &
Recommendations

Candidate Scoring

Shortlisting
Best-fit Recommendations

AI Automation

Workflow Automation

Scheduling Support
Recruitment Process Optimization

Insights

Hiring Analytics

Feedback + Ratings
Decision Support

Information architecture

Working my way down, thinking from a high level structure…

Proposed following design interventention

Recruiter

Dashboard view with major recruitment metric movers. Automation in time-intensive tasks such as JD creation, sourcing, screening, and interview scheduling to save time and decrease costs.

Hiring Manager

Leverage a candidate ranking and evaluation tool to evaluate candidate-job fit percentage and avoid mis-hires in the team.

Recruiter

Dashboard view with major recruitment metric movers. Automation in time-intensive tasks such as JD creation, sourcing, screening, and interview scheduling to save time and decrease costs.

Hiring Manager

Leverage a candidate ranking and evaluation tool to evaluate candidate-job fit percentage and avoid mis-hires in the team.

Wireframes

DESIGN SYSTEM & ACCESSIBILITY

🎨 Design System
(AIR Design system with AIR Icons )

Reusable components

Consistent typography & spacing
Standardized forms & tables
AI interaction patterns
Responsive layouts
Developer-friendly design tokens

SOLUTION IN DETAIL V1

DASHBOARD

JD CREATION + AI EDITOR & TAGGING FEATURE

SCREENING OF CANDIDATES & SUCCESS METRICS

SCHEDULING INTERVIEW FULLY CHAT BASED & ALLOCATING

THE INTERVIEWER & SHORLISTED CANDIDATES

FEEDBACK FORM OF CANDIDATES & RATING SECTIONS

USABILITY TESTING wit

I conducted moderated usability testing using interactive Figma prototypes with 10-12 participants, including recruiters, hiring managers, product managers, and internal stakeholders familiar with enterprise hiring workflows from Pearson, Select quote & Genpact.

Check Figma workflows

Tasks Tested

Create a Job Description using AI

Screen and shortlist candidates
Review AI match scores
Schedule interviews
Submit interview feedback

Method

Task-based testing
Prototype walkthroughs

Insights collected from the testing phase

Problem: Users overlooked the AI chat.
Solution: Made AI the primary entry point with quick action prompts.

Problem: Users saw AI scores but not the reasoning.
Solution: Displayed key match insights directly on candidate cards.

Problem: Blank prompts caused hesitation.
Solution: Suggested existing JD templates before AI generation.

Problem: Users hesitated to trust AI-generated feedback.
Solution: Clearly labeled feedback as AI-generated and editable.

Problem: Rankings lacked clear next steps.
Solution: Added contextual actions like Offer, Next Round, and Reject.

Problem: All the major navigations are at bottom. users felt empty around
Solution: Added a side left navigation panel with all features and made the search bar a seperate entity.

% Outcome from testing

100% Task Completion
After Round 1 improvements, all participants successfully completed the core tasks, validating the revised workflow.
86% Faster First Interaction
First meaningful interaction improved from 3m 40s to less then 30s.
40% Faster Feedback
Feedback submission time improved by 40%, with no reported hesitation around AI transparency.
100% Conversational Preference
100% of participants preferred the conversational experience over the traditional ATS workflow.
Commercial Validation
Following Round 2 improvements, the product secured its first commercial commitment.

FINAL DESIGN - IMPLEMENTED V2

Two Themes Implemented - Apricot, Blueberry

FINAL VERSION PROTOTYPE

Recruit Assist Live

KEY LEARNINGS

Built trust in AI through transparent, explainable, and human-in-the-loop experiences.
Validated design decisions with early usability testing and research-driven iterations.
Collaborated effectively with cross functional teams to align user, business, and technical goals.
Strengthened leadership through mentoring, constructive feedback, and continuous design audits.

Built trust in AI through transparent, explainable, and human-in-the-loop experiences.
Validated design decisions with early usability testing and research-driven iterations.
Collaborated effectively with cross functional teams to align user, business, and technical goals.
Strengthened leadership through mentoring, constructive feedback, and continuous design audits.

Built trust in AI through transparent, explainable, and human-in-the-loop experiences.
Validated design decisions with early usability testing and research-driven iterations.
Collaborated effectively with cross functional teams to align user, business, and technical goals.
Strengthened leadership through mentoring, constructive feedback, and continuous design audits.

FUTURE SCOPE

Agentic Recruitment - AI agents execute sourcing, screening, scheduling and follow-ups.
Design style - Will update and change the design style with more cool intercations & colors
Predictive Hiring - Forecast candidate fit, hiring bottlenecks and time-to-fill.
Continuous Matching - Continuously match candidates using skills and transferable experience.
Personalized Copilot - Adapt recommendations to each recruiter’s workflow and preferences.

THAT'S A WRAP, HOPE YOU LIKED IT!