About
NInsideN
Neural Inside Network
What NInsideN Means
NInsideN stands for Neural Inside Network. The name captures the core mission: to make what happens inside an AI system visible, understandable, and interactive — without mystifying or overpromising.
Modern AI products are powerful, but the process is largely hidden. NInsideN makes the observable, application-level pipeline transparent.
Mission
Make AI systems visible, understandable, and interactive. Not through hype or magic claims — through honest, technically accurate visualisation of real application-level processes.
What NInsideN Shows
- →How a prompt is received and normalised
- →How text is split into tokens
- →How intent is detected from the prompt
- →How a retrieval query is generated
- →How local demo knowledge is searched
- →How context chunks are selected and assembled
- →When and how the model request is made
- →How tokens stream back from the model
- →How the final answer is assembled
What NInsideN Does Not Claim
NInsideN does not claim to expose private hidden AI reasoning, chain-of-thought, or internal model weights.
The model is a black box. NInsideN makes visible the observable application pipeline that surrounds it: prompt processing, retrieval, context building, the model API request, and the streaming response.
Technology
Frontend
Next.js · TypeScript · Tailwind · Framer Motion
Backend
Python · FastAPI · SSE · OpenAI API
AI API
OpenAI-compatible chat completions streaming
Deployment
Vercel — ninsiden.com