How FYJ Works
What powers the platform: named research sources and named coaching frameworks, with every claim traceable to where it came from.
Research & data sources
Public data the AI draws on the moment a student starts a conversation. Federal data, occupation databases, and independent labor research.
Coaching frameworks the AI applies
Established frameworks from recruiting, leadership, and coaching literature. Each card cites its origin so students and reviewers can trace the method back to its source.
Principles that hold these together
When the AI surfaces a salary, it links to the federal data. When it describes a career path, it links to the occupation source. The design target is traceability for every number students see.
The AI reads the student's actual coursework, grades, experiences, and goals. Generic career advice is the baseline other tools offer. Grounding in the individual student is what makes FYJ different.
Source grounding and post-generation validation are in place to reduce the risk of invented facts. Like any AI system, occasional errors are possible. The architecture is designed to catch them.
What you can count on
Standards and commitments that apply to every FYJ engagement, regardless of audience.
- Your data stays yours. Each user's data is isolated. No cross-tenant access.
- No training on user data. User content is not used to train AI models. OpenAI API data is not used by OpenAI for model training per their enterprise API terms.
- Deletion on request. Any account or its data can be removed. Verified in writing.
- SOC 2 Type II infrastructure. Hosted on independently audited platforms (Supabase and AWS).
- Encryption everywhere. TLS 1.2+ in transit, AES-256 at rest.
- Built to WCAG 2.1 AA accessibility principles. Keyboard navigation, screen reader support, color contrast, and semantic structure throughout.
FYJ partners with universities to bring this foundation into their career centers, tailored to each school’s curriculum, voice, and systems.