Career OS
A structured career-management platform for tracking opportunities, applications, hiring workflows, professional evidence, and future AI-assisted career intelligence.
Origin
Career and job-search information is usually fragmented across spreadsheets, resumes, job descriptions, emails, interview notes, and recruiter conversations. None of it is structured, reusable, or connected.
Career OS consolidates that information into a structured career knowledge base: opportunities, applications, hiring timelines, professional evidence, and operational follow-ups — modeled as a production system rather than a demo CRUD. The long-term goal is a Career Knowledge Base that can ground future AI-assisted career workflows.
It is built as a production-oriented modular monolith, not a throwaway prototype. Domain modeling, temporal consistency, and a typed API contract are first-class concerns because the system is designed to hold real, correction-friendly historical data over time.
Implemented / Current
Opportunity Management
Structured capture of job opportunities with controlled lifecycle.
- Company, role, employment type, workplace, country/location, compensation, source
- Discovered/expiration dates, priority/interest, filtering
- List, detail, and edit flows
- Controlled lifecycle: open / pursuing / passed / closed
- Conversion of opportunities into applications
Application Tracking
Applications created from opportunities, with submission context and linked opportunity.
- Create application from an opportunity
- List, detail, and edit flows
- Submission information and linked opportunity context
- Contacts and participants foundations
Hiring Timeline
Application status and stage are derived from the event timeline rather than freely edited.
- Timeline events: Applied, Screening, Assessment, Interview, Offer, Rejected, Withdrawn, Hired, Other
- No rigid universal hiring-process state machine
- State derived from historical events
Actions & Expectations
Operational tracking distinguishing what the candidate must do from what they are waiting for.
- Action — something the candidate needs to do
- Expectation — something the candidate is waiting for
- Due/expected dates, completion/resolution dates, notes
- Pending / completed / cancelled state, linked to timeline events
Temporal Conflict Resolution
Historical date changes can affect dependent records. Career OS detects impact and corrects atomically.
- Detects downstream conflicts from date changes
- Returns structured conflicts
- Review Date Conflicts workflow
- Revalidates resulting state and applies changes atomically
Cancellation / Rescheduling
Schedulable events (Screening, Assessment, Interview) can be cancelled or rescheduled with provenance.
- Cancelled outcome and cancellation date
- Replacement events instead of silent overwrites
- Reschedule chains with one-step Undo Reschedule
- Chain-aware date constraints and protected chain deletion
- Compact provenance in the timeline
Structured API Validation
Backend errors follow a structured contract mapped to localized frontend messages.
- Structured errors: field, code, message
- Frontend field-level and global form errors
- Localized domain messages
Career & Job Intelligence Foundations
Structured foundations for professional experience, projects, technologies, achievements, and job intelligence data.
- Foundations for a structured Career Knowledge Base
- Backend foundations for job descriptions / job intelligence data
- Designed as future AI context, not as a shipped AI platform
Architecture
Supporting
Engineering Decisions
Timeline as source of truth
Application state is derived from an authoritative event timeline rather than manually synchronized fields. This keeps status, stage, and history consistent and auditable instead of relying on hand-edited state.
Temporal conflict resolution
Changing an application or event date can affect dependent records. Career OS detects downstream impact, lets the user review corrections, and applies the final state atomically — so historical edits cannot silently corrupt the timeline.
Correction-friendly historical data
The system supports backfilled histories and corrections instead of assuming all data is entered live. Real application histories are often reconstructed after the fact, and the model accommodates that without losing consistency.
Reschedule provenance
A rescheduled interview preserves the original event and creates a linked replacement instead of silently overwriting history. Reschedule chains remain auditable and support one-step undo with chain-aware date constraints.
Typed API contract
FastAPI/OpenAPI-generated frontend types plus structured AppError mapping keep the React and Python halves in sync. Field-level, code-driven errors flow into localized frontend messages instead of opaque strings.
Same-day event sequencing
Timeline ordering does not depend on when a record was created. occurred_at captures when something happened, same-day sequence captures order within that date, and created_at only records when Career OS logged it.
Quality
- Pytest for backend tests
- Vitest for frontend tests
- Playwright for end-to-end tests
- Alembic migration tests
- Type checking and linting (Ruff)
- OpenAPI contract generation
- Manual dogfooding with real application histories
AI-Assisted Engineering Workflow
Career OS is developed using a spec-driven, AI-assisted engineering workflow. AI agents accelerate implementation while engineering ownership stays human.
Human ownership
- Product decisions and domain modeling
- Architecture, constraints, and tradeoffs
- Reviewing AI output
- Manual testing and validation
- Deciding when work is complete
- Documenting durable decisions in ADRs
AI assists
- Codebase analysis and investigation
- Implementation and refactoring
- Tests and migrations
- Debugging and edge-case discovery
Planned AI Capabilities
Future direction — not shipped. Career OS is architected toward AI-assisted workflows grounded in structured career data, but no AI integration is currently implemented.
Career Knowledge Base as AI context
Future AI workflows intended to use structured evidence — experience, projects, achievements, technologies, job descriptions, applications, and interview history.
Job / Opportunity Intelligence
Potential future workflows: job-description analysis, job/profile comparison, skills and gap identification, preparation recommendations, and career-goal alignment.
Contextual career documents
Future direction: evidence-grounded generation of tailored CVs/resumes, cover letters, interview preparation, STAR stories, and role-specific materials.
Interview Intelligence
Planned concepts: interview stages, technical and system-design topics, behavioral questions, lessons learned, and preparation packs for similar future processes.
Career Direction & Gap Analysis
Future backlog: target role, compensation goals, IC vs management, specialization, strengths and gaps, evidence gaps, and recommended projects or learning priorities.
Learning & AI project recommendations
Future direction: prioritized learning plans and project recommendations based on real gaps, repeated technologies from job descriptions, and portfolio evidence gaps.