Working product: chat assistant, browser companion, and custom RAG chatbots trained on your documents. Running on optimised open-weight models.
AskBurg AI Systems · [email protected]
AskBurg is a working product you can use today — not just a landing page.
Real-time streaming chat with conversation memory. Scroll down to try it — no signup required for the demo.
SaaS dashboard for document ingestion, RAG search testing, and API key management. Custom chatbots trained on your data.
Sidebar chat, page summaries, selection toolbar, and compose assistant. Works in Chrome, Firefox, Edge, and Brave.
Model Strategy: Open-weight foundation models adapted with LoRA
fine-tuning and optimised with AWQ and GGUF quantization, including for cost-sensitive
and air-gapped deployments.
Stack: Node.js/Express backend,
vanilla JS frontend (no build step), in-memory RAG with citation tracking.
Deploy a 24/7 web chatbot trained on your documentation, FAQs, and product manuals to answer customers instantly.
Ground answers directly in internal company documents with verifiable source citations and strict data isolation.
Summarize long documents, explain complex web articles, and draft polished emails directly within your active browser tab.
Production-ready features built into the platform.
Server-Sent Events (SSE) deliver answers word-by-word. No waiting for complete responses.
Conversation history lets you ask follow-up questions naturally, with full context.
Upload documents, chunk content, and get citation-grounded answers from your private knowledge base.
REST endpoints with API key auth, rate limiting, and streaming SSE support for integration.
Type any question in the chat interface — from technical explanations to writing help.
Your query runs through optimised open-weight models with strict privacy controls.
Watch the response appear word-by-word in real time. Continue the conversation naturally.
Beyond the public assistant, deploy private RAG chatbots trained on your documents with verifiable citations.
Upload PDFs, docs, and knowledge base articles. Automatic chunking and keyword indexing for retrieval.
Every answer includes bracketed source citations ([Doc Title, Section N]) so you can verify claims.
Deploy chatbots as web widgets or REST APIs. White-label options available for enterprise customers.
Your documents stay private, never used for training. 30-day retention policy, opt-in model training only.
Live demo — no signup required. Ask anything and see real-time streaming responses.
You're chatting with AskBurg, an AI assistant, not a human. AskBurg can make mistakes. Double-check important info. Not legal, medical, or financial advice.