ORIEN Intelligence Engine

ORIEN is not a traditional assessment engine. It is an Industry Intelligence Engine that continuously evolves to understand modern roles, competencies, and business contexts before generating assessments.

The Core Equation

The entire SMART platform relies on the premise that competencies must be generated, not merely selected from a static list.

Technical Implementation (PostgreSQL + JSONB)

ORIEN is powered by PostgreSQL coupled with pgvector and Prisma ORM.

• Why JSONB? The Competency Blueprint is a highly nested, evolving AI output. Using JSONB allows ORIEN to generate flexible schemas while maintaining transactional integrity in the larger relational database (for Orgs, Users, and RBAC).
• Explainable AI: Every competency recommendation stores a reasoningTrail referencing the sources (e.g., “Suggested System Design because 84% of Senior Backend roles in Fintech require distributed systems knowledge”).
• Anthropic Claude Integration: We utilize Anthropic Claude (Sonnet) via the ai SDK, orchestrated through BullMQ/Redis for asynchronous, reliable background processing.

Continuous Knowledge Graph

ORIEN builds its intelligence by scraping and ingesting real-world data:

1. Public JDs & Career Pages: Understanding what the market currently demands.
2. Engineering Blogs: Extracting modern technology stacks and patterns.
3. LinkedIn & Github Signals: Mapping skill correlations.
4. Funding & Market Trends: Adapting requirements based on company stage.