Opportunities & Applications
Organize opportunities, companies, job descriptions, application stages, status and the overall application lifecycle.
Featured Technology Project • Live
Job Search
Job Search
Manager
Simple by Design. Secure by Default.
A secure, cloud-native application created from a real-world need to bring opportunities, relationships, career documents and job-search activities into one structured system.
Cloud-Native
Security by Design
Full-Stack
AI-Assisted Engineering
01 • The Origin
Why I Built It
Job Search Manager began with a problem I experienced personally while managing an active and increasingly complex job search.
Applying for a serious opportunity involved much more than finding a job and submitting a resume. Each role required reviewing the Job Description, assessing how well my experience matched, evaluating ATS alignment, identifying relevant skills and keywords, customizing the resume where appropriate, and tracking exactly which resume version was submitted.
At the same time, I needed to manage opportunities from different channels, recruiter and professional contacts, application status, interviews, follow-ups, job descriptions, resumes and supporting documents. The workflow quickly became fragmented across spreadsheets, job portals, documents, email, LinkedIn and other channels.
Rather than continuing to adapt my workflow around disconnected tools, I decided to design and build a purpose-built system. What began as a resume-management idea evolved into an end-to-end Job Search Manager.
The Real Workflow
01Review the Job Description
02Assess Match & ATS Alignment
03Customize & Version the Resume
04Apply & Track
05Connect & Follow Up
06Interview & Manage Outcome
The goal was not to over-engineer the solution. Simplicity would never come at the expense of security, privacy, maintainability or engineering discipline.
Simple by Design. Secure by Default.
02 • Product
What the Product Brings Together
A single structured system supporting the broader job-search lifecycle rather than one isolated task.
Contacts & Relationships
Bring recruiters, hiring contacts, professional connections, interactions and follow-up context into the same workflow.
Resumes & Documents
Maintain career assets, resume versions and supporting documents so the right material can be associated with the right opportunity.
JD & Application Workflow
Keep the job description and application context connected, supporting a structured approach to evaluating and preparing for opportunities.
Interviews & Follow-Ups
Keep interview and follow-up activity connected to the broader opportunity journey rather than scattered across separate notes and channels.
Dashboard & Visibility
Create a consolidated view of job-search activity and pipeline information to make the overall process easier to understand and manage.
Administration
Support controlled administration and user-management capabilities with protected privileged operations and ownership rules.
Security & Access
Apply authenticated access, role-aware authorization, session controls and protected API boundaries across the application.
03 • Product Map
The Job-Search Journey
The product is organized around the journey of the job seeker—not around a collection of disconnected software modules.
01Discover &
EvaluateOpportunities • JDs
EvaluateOpportunities • JDs
→
02Prepare &
TailorResumes • Documents
TailorResumes • Documents
→
03Apply &
TrackApplications • Status
TrackApplications • Status
→
04Connect &
Follow UpContacts • Interactions
Follow UpContacts • Interactions
→
05InterviewInterviews • Follow-ups
→
06OutcomePipeline • Visibility
Evolution Roadmap
Deeper ATS analysis, resume editing and customization, job feeds and matching, and mock interview/coaching capabilities are treated as product evolution—not presented here as completed production capabilities.
04 • Architecture
Cloudflare-First. Purposefully Simple.
The architecture was selected to fit the problem: secure, maintainable, cost-conscious and scalable enough for its purpose without unnecessary infrastructure.
USERBrowser / Web UI
↓
CLOUDFLARE PLATFORM / APPLICATION BOUNDARY
FRONTENDReact + ViteApplication Experience
PROTECTED BACKENDCloudflare WorkersAPI • Business Logic • Security Enforcement
Authentication • Authorization • Session Controls • Validation • Business Rules
STRUCTURED DATACloudflare D1Application Records
OBJECT STORAGECloudflare R2Documents & Files
Protected API Boundary
Privileged operations are mediated through the backend/API layer rather than giving the browser direct privileged access to data and storage resources.
Separation of Concerns
D1 holds structured application records while R2 handles file/object storage—using each service for the type of data it is designed to manage.
Appropriate Complexity
The design deliberately avoids infrastructure that does not add meaningful value to the problem, reducing operational burden without removing necessary controls.
05 • Security by Design
Security Was a Requirement, Not an Afterthought
The application handles personal career information and documents. Appropriate controls were therefore considered across identity, sessions, authorization, APIs, data, development and deployment.
Mandatory MFA for Administrators
Privileged administrative access requires stronger authentication, applying a simple principle: higher privilege requires stronger identity assurance.
Server-Enforced Session Timeout
Authenticated sessions are subject to server-side timeout controls so unattended access does not remain available indefinitely.
Server-Side Authorization
Role and privilege controls are enforced by the backend rather than relying only on what the browser chooses to display.
Protected Data Access
Application data and files sit behind controlled application/API boundaries instead of being exposed as directly privileged browser resources.
01Identity & AuthenticationAuthenticated access • stronger controls for privileged administration
02Session SecurityServer-side validation • inactivity / expiry controls
03Authorization & OwnershipRBAC • protected owner/admin operations • data ownership concepts
04API & Application ControlsProtected endpoints • validation • business-rule enforcement
05Data & Storage BoundariesStructured data and file storage separated behind controlled services
06Engineering & Change ControlBranching • review • testing • controlled integration
Some of the Most Important Features Are Invisible
Authentication controls • authorization checks • ownership validation • protected APIs • session handling • privileged-access safeguards • data boundaries • AI guardrails • code review • Git change control
06 • Technology & Local Development
Technology Chosen to Fit the Problem
The stack was selected around security, simplicity, maintainability, cost efficiency and the ability to evolve—not around collecting technologies.
Responsive application UI and modern local development workflow.
Protected server-side API, business logic and security enforcement.
Relational storage for structured application records.
Separate storage boundary for documents and file assets.
Local Worker runtime, bindings, configuration, migrations and deployment workflow.
Development, dependency and application build tooling.
Branching, traceability, known-good checkpoints and controlled integration.
UI, network/API, authentication, cookie and integration troubleshooting.
Local Development Environment
Build and Validate Before Production
The frontend, backend/API, database migrations and service bindings were exercised through a local development workflow designed to represent the deployed architecture as closely as practical.
React / ViteWrangler / WorkerD1 / MigrationsR2 BindingsBrowser DevToolsGit
07 • Git, Branching & Change Control
Isolate the Change. Protect the Baseline.
Git was used as an engineering control—not simply as a place to store finished code.
mainStable Integrated Baseline
task-x.x-feature-nameDefined / Isolated Change
Implement→Build & Test→Review→Correct→Known-Good Commit
↓ Controlled Merge ↓
mainValidated Baseline → Deploy
Task-Oriented Branches
Meaningful capabilities and changes were isolated in dedicated branches so development could progress without immediately putting the stable baseline at risk.
Known-Good Checkpoints
Commits were used to preserve validated states, providing traceability, recoverability and safe points before moving to the next task.
Controlled Integration
Changes moved toward the stable branch after implementation, local validation, review and correction rather than through uncontrolled large-scale edits.
08 • AI-Assisted Engineering
AI as an Accelerator—Inside Engineering Boundaries
AI accelerated implementation, exploration, troubleshooting and review, but generated output was not automatically treated as production-ready.
“AI can accelerate development. Engineering discipline provides the control.”
01Human DirectionProblem • Requirements • SRS • Decisions
→
02ArchitectureDesign • Security • Boundaries
→
03AI-Assisted BuildImplementation • Exploration • Debugging
→
04Review & TestFunctional • Architecture • Security
→
05Git ControlCommit • Merge • Traceability
→
06Deploy & ValidateProduction verification
Gemini
Used substantially for larger implementation and code-generation tasks within defined requirements and task boundaries.
ChatGPT
Used for requirements refinement, architecture discussions, implementation planning, debugging, change review, security reasoning and iterative refinement.
AI Guardrails
AI as a Controlled Capability
AI capabilities were designed to operate within the application's existing security and engineering boundaries—not as a shortcut around them. AI-related access remains subject to authenticated application flows, server-side controls and defined task boundaries, while consequential decisions remain with the user.
Authenticated Access
Server-Side Control
Defined Scope
Controlled Context
User Review
No Autonomous Privileged Action
Guardrail Principle
AI is a capability—not a trust boundary. Authentication, authorization, data boundaries and application controls continue to govern how AI-enabled functionality is accessed and used.
09 • Code Review & Quality
Working Code Was Not Automatically Finished Code
Review looked beyond whether a page loaded. Changes were challenged for correctness, architecture, security, scope and regression risk.
Does it satisfy the requirement?
Validate expected workflows, behavior and whether the implementation actually solves the defined task.
Does it still fit the design?
Check separation of concerns, service boundaries, unnecessary complexity and architecture drift.
Are controls enforced correctly?
Review authentication, authorization, privileged operations, session behavior, ownership and protected APIs.
What should not have changed?
Look for unintended file changes, broken existing behavior and side effects outside the intended scope.
Review principle: Verify not only what changed—but also what should not have changed.
10 • Testing & Security Validation
Test Expected Success—and Expected Denial
Testing was incremental and layered, covering application behavior, integration, authentication/authorization and security-control outcomes.
Local Functional Testing→API / Integration→Auth & Authorization→Security Controls→Regression Checks→Production Validation
Positive Paths
Confirm that authorized users and valid workflows behave as expected.
Negative Paths
Confirm that unauthenticated, insufficiently privileged or expired access is denied where it should be.
Real Integration Issues
Development included resolving practical issues involving CORS, authentication state, session cookies, protected routes, API authorization, migrations and deployment behavior.
11 • Delivery
From Defined Task to Production
Task Defined→Task Branch→AI-Assisted Implementation→Local Development→Code & Security Review→Test & Retest→Known-Good Commit→Controlled Merge→Deploy→Production Validation
The objective was not process for the sake of process. Each control created a safer boundary around rapid, AI-assisted development.
12 • Outcome & Evolution
From Personal Problem to Deployed Product
The achievement is not only the application itself—it is the complete journey from identifying a real problem to designing, securing, reviewing, testing and deploying a working solution.
Product Thinking
Started with a real operational problem and expanded the scope only when the broader lifecycle made the need clear.
Architecture Thinking
Selected a pragmatic cloud architecture that balanced security, simplicity, maintainability and cost.
Security Thinking
Treated identity, privileged access, sessions, authorization and data boundaries as design requirements rather than late additions.
Engineering Discipline
Used local environments, task-based branching, reviews, testing, known-good checkpoints and controlled integration around AI-assisted development.
Product Evolution
What Comes Next
The roadmap extends the same end-to-end job-search vision with deeper intelligence and assistance while preserving the principles established in the first production foundation.
Resume Editing & Tailoring
Deeper ATS / JD Analysis
Job Feeds & Match Intelligence
Mock Interviews & Coaching
Job Search Manager
Real Problem. Thoughtful Design.
Real Problem. Thoughtful Design.
Working Product.
Built to solve a problem I experienced firsthand—and developed as a practical exercise in product thinking, secure architecture, modern cloud engineering, disciplined change control and responsible AI-assisted development.