Work
Ife - Internal Coding Intelligence
Overview
Ife is Blacksky's internal coding intelligence system - a private RAG-based assistant that provides Claude/Gemini-level guidance while keeping proprietary code secure within Blacksky's infrastructure. Ife knows Blacksky's codebase, conventions, and architecture, helping developers without ever sending proprietary code to external APIs.
The Name
Ife (ee-feh) is Yoruba for "clarity." The name reflects Blacksky's commitment to non-European narratives in technology.
The philosophical foundation comes from Malcolm X's 1964 Oxford Union speech about breaking from restrictive systems. Applied to Ife: breaking from "black-box" public models toward open, collaborative, clear internal tooling. Data sovereignty as a technical and cultural value.
The Problem
Public AI coding assistants like GitHub Copilot and ChatGPT require sending your code to external servers. For proprietary work, that's a problem:
- Client code could be exposed to third-party training data
- Compliance requirements may prohibit external data sharing
- No guarantee of data deletion or privacy
- Dependency on external API availability and pricing
Ife provides the same intelligence level while keeping code private - productivity without compromise.
What Ife Does
- Contextual coding assistance trained on Blacksky's actual codebase
- Institutional knowledge about documentation, conventions, and best practices
- Project intelligence that understands internal architecture and patterns
- Developer support without external API calls for core functionality
Ife isn't a chatbot - it's a senior knowledge architect that understands Blacksky's specific context.
Technical Architecture
| Component | Technology | |-----------|------------| | LLM | Local deployment (Llama-based) | | RAG | Custom pipeline on Blacksky codebase | | Vector Database | ChromaDB | | Backend | Python |
The RAG pipeline continuously indexes Blacksky's repositories, documentation, and internal resources. When a developer asks a question, Ife retrieves relevant context from the codebase before generating a response - answers grounded in actual Blacksky code, not generic patterns.
Key Differentiators
- Private - Code never leaves Blacksky infrastructure
- Contextual - Knows our stack (FastAPI, Next.js, Drupal, Python, etc.)
- Current - RAG updates as codebase evolves
- Secure - No external API dependencies for core functionality
- Culturally grounded - Named and designed with intention
Why This Project Matters
Ife demonstrates Blacksky's approach to AI tooling:
- Build, don't just buy - Instead of depending on external AI services, build internal capabilities
- Data sovereignty - Keep proprietary information private while still leveraging AI
- Contextual intelligence - Generic AI is useful; AI that knows your specific codebase is transformative
- Cultural intention - Technology reflects values, including whose narratives are centered
For clients concerned about AI and data privacy, Ife proves Blacksky understands the challenge - because we solved it for ourselves first.
Project Type: Internal Tool Status: In development - internal use at Blacksky Deployment: Local infrastructure