RaderonLab AI Core
The shared AI engine behind the lab's products.
The problem it solved
Every product was re-implementing AI plumbing (OCR, LLM calls, tools) from scratch. This consolidates it into one reusable AI engine the whole portfolio plugs into.
Similar problems it can solve
A shared AI backbone for any product suite: OCR/LLM/tool-orchestration as a service, rapid AI-feature bootstrapping, and vendor-agnostic model integration.
A reusable AI middleware service that several products plug into as their AI core — OCR, LLM orchestration and external-tool integration, exposed through a web UI and a service API on .NET 10.
- Shared OCR + LLM-orchestration core
- Consumed by multiple products as the AI engine
- Web UI + service API + external-tool integrations
Overview
RaderonLab AI Core is the shared AI engine that the lab's products plug into instead of each re-building AI from scratch — a single, reusable middleware for OCR, LLM orchestration and tool use.
One core, many consumers
Around the core sit OCR test harnesses (console, ASP.NET Core and WPF), a web UI, an external-tools integration service and a research module — and downstream products (document intelligence, translation engines, trading tooling) consume it as their AI backbone.
Engineering depth
It abstracts vision-language OCR (DeepSeek-OCR, Qwen2-VL, Gemini Flash post-processing) and LLM orchestration behind one service API, so swapping a model or adding a tool is a configuration change, not a rewrite — vendor-agnostic by design.
Technology
.NET 10, ASP.NET Core, vision-language models, LLM orchestration (Gemini and others), and a service-API + web-UI surface.