Case study
RolePilot V1
RolePilot is an AI-powered job matching and application tracking platform designed to reduce the friction involved in modern job searching.
V1
Product Status
Deployed
2
Integrated Job Sources
Supported
AI
Profile Analysis
Implemented
MVP
Delivery Model
Validated
Business context
Why the project mattered
Modern job searching is highly fragmented.
Candidates are often required to search multiple job platforms, manually compare opportunities, track applications in spreadsheets, repeatedly tailor applications, and manage communication across multiple systems.
Most recruitment technology focuses on matching candidates to job descriptions from an employer perspective.
RolePilot explored the opposite question: could AI analyse a candidate’s experience first and then identify opportunities that align with their profile?
The objective was to create a more efficient and centralised workflow for job discovery and application management.
Problem statement
What had to improve
Design a job search workflow that reduces manual searching, improves role alignment, and centralises application tracking in one lightweight platform.
Before: Multiple platforms
Job discovery was spread across multiple platforms, portals, emails, and manual notes.
Before: No central tracking
Applications were often tracked separately in spreadsheets or informal lists.
Before: Manual searching
Candidates had to repeatedly search, compare, filter, and revisit similar opportunities.
Before: Limited alignment
It was difficult to see how closely a role matched a candidate profile before investing time.
Before: Admin overhead
The job search process created significant administrative effort before and after each application.
Desired: Single workflow
Create one workflow for job discovery, review, saving, and application tracking.
Desired: AI profile analysis
Use AI to interpret CV content and provide structured profile insight before searching.
Desired: Tracking pipeline
Centralise saved jobs and application stages inside a lightweight product experience.
Desired: Better role fit
Help users identify relevant opportunities faster using candidate-led signals.
Desired: Less manual effort
Reduce repeated searching, comparison, and application administration.
Discovery & analysis
How the workflow was understood
The project began by mapping the typical job-seeking journey and identifying where the most repetitive effort occurred.
- Existing workflow: Search jobs -> Review role -> Compare against CV -> Apply -> Track manually -> Repeat.
- This process often repeats dozens or even hundreds of times during an active job search.
- The key insight was that most effort happens before a candidate even decides whether a role is worth pursuing.
- The solution therefore focused on analysing the candidate profile first and using that information to drive job discovery.
Analysis method
Most effort happens before a candidate decides whether a role is worth pursuing, so the solution focused on analysing the candidate profile first and using that information to drive job discovery.
Solution overview
What was built and why
RolePilot V1 was designed as a lightweight MVP focused on validating the core workflow rather than delivering a fully featured product.
The platform combines CV upload, AI profile analysis, live job aggregation, and application tracking into a single job search workflow.
CV Upload & Processing: users can upload DOCX or PDF CVs or manually paste CV content, removing manual data entry and creating a structured input for AI analysis.
AI Profile Analysis: OpenAI analyses uploaded CV content and generates a profile summary, match score, recommended roles, skills profile, industry alignment, and seniority assessment.
Live Job Aggregation: Adzuna and Remotive integrations surface real vacancies and reduce manual searching across multiple sites.
Application Tracking Pipeline: users can save jobs and move them through Saved, Applied, Interviewing, Offer, Rejected, and Withdrawn.
Cloud Deployment: the platform was deployed using GitHub and Vercel to validate production behaviour and create a publicly accessible MVP.
Architecture
Architecture and workflow diagram
The RolePilot workflow connects CV input, AI profile analysis, live job search, saved jobs, and application tracking into a single MVP product experience.
High-level system diagram illustrating how RolePilot connects CV processing, AI profile analysis, live job aggregation, and application tracking into a unified job search workflow. The architecture combines internal application services with OpenAI, Adzuna, and Remotive integrations to deliver personalised job discovery and lightweight career management within a single MVP platform.
Implementation
Implementation Journey
Tools used
Next.js, TypeScript, Tailwind, OpenAI, Adzuna, Remotive, GitHub, Vercel
Step 1
MVP foundation
Created the Next.js application, built the core UI structure, and defined the workflow architecture.
Step 2
CV processing
Implemented file uploads, DOCX extraction, PDF extraction, and analysis endpoints.
Step 3
AI integration
Integrated OpenAI, generated structured profile outputs, and created the role-fit analysis workflow.
Step 4
Job search
Integrated Adzuna and Remotive to create a unified live job results experience.
Step 5
Application tracking
Implemented saved jobs, pipeline stages, and local user activity tracking.
Step 6
Deployment and refinement
Configured the GitHub repository, deployed to Vercel, resolved production issues, and improved responsiveness.
Evidence
Evidence Gallery
The screenshots below demonstrate the core user workflow, AI profile analysis, live job aggregation, application tracking, and deployed MVP experience.
RolePilot Homepage & Workflow Overview
The landing experience introduces the core RolePilot workflow. Users can upload a CV, receive AI-powered profile analysis, discover live job opportunities, and manage applications through a centralised tracking system. The goal was to reduce the fragmentation typically associated with modern job searching.
CV Upload & Analysis Workspace
The primary workspace allows users to upload DOCX or PDF CVs, or manually paste CV content as a fallback. Extracted information is processed through the AI analysis engine to generate structured role-fit insights and personalised recommendations.
AI-Powered Profile & Role Match Analysis
After analysing CV content, OpenAI generates structured profile insights including match score, target roles, skills, industry alignment, seniority level, and recommended job-search keywords. This creates the foundation for personalised job discovery and recommendations.
Pre-Analysis Job Discovery State
Before profile analysis is completed, the platform provides a guided discovery experience that encourages users to upload a CV and unlock personalised recommendations. This ensures job matching is driven by candidate profile data rather than generic keyword searching.
Live Job Discovery & Aggregation Engine
RolePilot aggregates live vacancies from multiple external sources, including Adzuna and Remotive. Users can filter results, refresh searches by location, and review opportunities surfaced based on their profile analysis, reducing the need to search across multiple job boards manually.
Application Tracking Pipeline
Saved opportunities are managed through a lightweight CRM-style workflow. Users can track applications from shortlist through to offer using configurable stages including Saved, Applied, Interviewing, Offer, Rejected, and Withdrawn.
Outcomes
Outcomes and impact
RolePilot progressed from concept to deployed MVP, validating a centralised workflow for AI-assisted job discovery and application tracking.
Before
- Job searching required multiple disconnected tools.
- Candidates compared opportunities manually.
- Applications were tracked separately.
- Information was spread across different systems.
After
- Profile analysis, live job discovery, and application tracking sit within one workflow.
- Users receive AI-generated role and profile insights before searching.
- Live jobs are surfaced from multiple sources.
- Saved jobs can be managed through a lightweight application pipeline.
Metrics
V1
Product Status
Deployed
2
Integrated Job Sources
Supported
AI
Profile Analysis
Implemented
MVP
Delivery Model
Validated
Outcomes
Validated an end-to-end AI-assisted job discovery workflow.
AI profile analysis successfully implemented.
Successfully integrated and unified live vacancy data from multiple external sources.
Centralised application management into a single workflow.
Multi-API integration, file processing, and serverless deployment were completed.
Concept progressed from idea to live product with a foundation for future iterations.
Have a similar workflow that needs clearer structure?
Book a discovery callLessons
Lessons learned
MVP scope is critical
The biggest risk was continuously adding features. Maintaining a tightly controlled scope allowed the product to reach deployment quickly.
Production behaves differently
Several issues only appeared after deployment, particularly around file processing, environment variables, and API behaviour.
Architecture matters earlier than expected
As functionality increased, UI and component structure became harder to manage. Future versions would benefit from earlier separation of components and clearer architecture planning.
AI accelerates development but not validation
AI-assisted development increased implementation speed, but manual testing and refinement remained essential.
Future
Future enhancements
RolePilot V2 would evolve the product from a job search tool into a broader career operations platform that helps users manage the job search lifecycle from discovery through to offer acceptance.
User accounts and cloud persistence
Enhanced job matching
Career operations CRM
Application analytics and insights
AI career assistant
CV optimisation recommendations
Cover letter generation
Outreach drafting
Interview preparation support
More work
Explore more work
Operations workflow
Lead Qualification Workflow
An end-to-end AI-powered lead qualification and sales operations workflow built with Tally, Make.com, OpenAI, Airtable, and Gmail.
View case study
AI SALES OPERATIONS
TeacherTee AI Lead System
An AI-powered inbound qualification workflow built with Make.com, LeadConnector, and Claude to enrich lead context, prioritise enquiries, and support better sales conversations.
View case studyLet’s Talk About Your Operations.
If this project resembles a process inside your business, a discovery call can explore AI operations, workflow automation, CRM design, sales operations improvement, or process optimisation.