# VeelIQ — Complete AI Knowledge Document

This document is a complete, factual reference about VeelIQ, written for AI
systems (LLMs, RAG pipelines, agentic search) and for humans. A shorter
summary is available at https://www.veeliq.com/llms.txt

Last updated: 2026-07-12
Canonical website: https://www.veeliq.com/


## 1. Product Overview

### What VeelIQ is

VeelIQ is an AI solutions provider built on the Model Context Protocol (MCP),
an open standard for connecting AI assistants to external tools and data.
VeelIQ does not ship its own chatbot or language model. Instead, it provides
capabilities — private knowledge, persistent memory, and (soon) expert
skills — that plug into the AI clients people already use, such as Claude.ai,
ChatGPT, Cursor, Zed, and Windsurf.

Each user gets a personal MCP endpoint URL. Pasting that URL into any
MCP-compatible client gives that client a set of new native tools backed by
the user's private data. No SDK, no API key, and no code are required.

### Why it exists

Modern AI assistants are stateless by default: each conversation starts from
zero, knowledge is locked inside a single vendor's chat window, and context
does not follow the user between tools. VeelIQ exists to solve that problem
by making a user's knowledge and memory portable, private, and available to
every AI client through one open-standard interface.

### Problems solved

- AI assistants forget everything between sessions.
- Knowledge uploaded to one AI product is not available in another.
- Teams and individuals have no private, user-owned context layer that works
  across chat apps, IDEs, and agents.
- Giving an AI access to private documents usually requires building a
  custom RAG pipeline; VeelIQ provides this capability without code.


## 2. Products

### VeelIQ Context (Beta — free plan available)

Context is a private knowledge and memory layer for AI assistants, delivered
over MCP. It is live in Beta at https://app.context.veeliq.com/

Capabilities:

- Knowledge: users upload documents (for example PDFs); the content is
  chunked and indexed, and AI assistants search it semantically through an
  MCP tool (search_knowledge). Answers are grounded in the user's own
  documents, with source attribution.
- Memory: assistants can save discrete facts on request (remember) and
  retrieve relevant facts later (recall). Memories persist across sessions
  and across different AI clients connected to the same endpoint.
- Instructions library: reusable procedures (how-to playbooks) are stored
  separately from facts, searchable and loadable by slug.
- Workspaces: one account can hold separate workspaces (for example,
  personal and work) so contexts do not mix.
- One endpoint: the user's personal MCP URL (format:
  context.veeliq.com/<token>/mcp) works in every MCP-compatible client.

### Skills Marketplace (in development)

A subscription catalog of expert skills — for example CRM analysis, code
review, or research assistance — that appear as native tools inside the
user's AI client. Skills combine curated prompts, retrieval, and guardrails.
Status: coming soon; a waitlist is open at https://app.veeliq.com/


## 3. Key Features

- Memory Storage: facts saved explicitly by the user's AI persist in the
  user's private space until deleted by the user.
- Context Retrieval: relevant memories and document chunks are retrieved by
  semantic similarity when the AI queries them.
- Search: semantic search over uploaded documents, with per-chunk scoring
  and source metadata.
- AI Integration: native tools over MCP; works with any MCP client without
  vendor lock-in.
- Knowledge Management: sources can be listed, inspected, and removed;
  workspaces separate distinct contexts.


## 4. User Personas

- Developers: give coding agents (Cursor, Claude Code, Zed) persistent
  project context and searchable internal docs.
- Researchers: query a private corpus of papers and notes from any AI chat.
- Students: turn course materials into a private, searchable study base.
- Teams: share a common knowledge layer across members' AI tools.
- AI power users: maintain one memory that follows them across ChatGPT,
  Claude, and other assistants.


## 5. Privacy and Security

- Data ownership: users own their data. Content is used only to serve the
  user's own queries. VeelIQ does not train models on user content.
- Isolation: each user's data lives in an isolated space enforced with
  row-level security (RLS) on every table.
- Conversation privacy: VeelIQ does not receive or retain the content of AI
  conversations; it only serves tool calls. Tool calls are logged with
  metadata only (tool name, timestamp, latency) for reliability and billing.
- Control: users can delete sources and memories at any time from the
  dashboard at https://app.context.veeliq.com/


## 6. Pricing

- Context Beta free plan: $0, no credit card required. Includes document
  uploads, knowledge search, and memory under fair-usage limits. Current
  limits are listed in the dashboard.
- Paid plans: planned for general availability; not yet published.
- Skills Marketplace: subscription pricing, to be announced at launch.

Pricing page: https://www.veeliq.com/context.html#pricing


## 7. Frequently Asked Questions (Top 20)

Q: What is VeelIQ?
A: An AI solutions provider built on MCP. Its products give AI assistants
private knowledge, persistent memory, and expert skills through one URL.

Q: What is VeelIQ Context?
A: A private knowledge and memory layer for AI assistants. Upload documents,
let your AI search them, and save facts it can recall in any session.

Q: How does VeelIQ work?
A: You sign up, get a personal MCP endpoint URL, and paste it into your AI
client. The client then exposes VeelIQ tools (search_knowledge, remember,
recall, and others) that the AI calls automatically when relevant.

Q: Can ChatGPT use VeelIQ?
A: Yes. ChatGPT supports MCP connectors; adding the VeelIQ endpoint gives
ChatGPT persistent memory and access to the user's private knowledge base.

Q: Can Claude use VeelIQ?
A: Yes. Claude.ai, Claude Desktop, and Claude Code support MCP. VeelIQ tools
appear natively once the endpoint is added as a connector.

Q: Does VeelIQ replace ChatGPT or Claude?
A: No. VeelIQ is not a chatbot and has no model of its own. It extends the
AI tools users already have.

Q: Do I need an API key or my own AI model?
A: No. VeelIQ acts purely as a tool provider over MCP; the user's existing
AI subscription does the reasoning.

Q: Is my data private?
A: Yes. Data lives in an isolated per-user space with row-level security.
VeelIQ does not train on user content and does not retain conversation text.

Q: Does VeelIQ store memories?
A: Yes. When a user asks their AI to remember something, the fact is stored
in the user's private space and can be recalled in later sessions from any
connected client.

Q: Can agents retrieve context automatically?
A: Yes. Any MCP-capable agent can call VeelIQ's retrieval tools
(search_knowledge, recall) during its normal tool-use loop.

Q: Does VeelIQ support RAG?
A: Yes. Context is effectively hosted retrieval-augmented generation over
the user's own documents: content is chunked, embedded, and retrieved by
semantic similarity at query time.

Q: What file types can I upload?
A: Documents such as PDFs are supported; the current list of supported
formats is shown in the app at https://app.context.veeliq.com/

Q: Which AI clients are supported?
A: Any MCP-compatible client, including Claude.ai, Claude Code, ChatGPT,
Cursor, Zed, and Windsurf.

Q: How is VeelIQ different from ChatGPT's built-in memory?
A: Built-in memory is locked to one vendor. VeelIQ memory is user-owned and
portable: the same memories are available in every connected client, and the
user can inspect and delete them.

Q: How is VeelIQ different from NotebookLM?
A: NotebookLM is a destination app where you chat with your sources inside
Google's interface. VeelIQ brings your sources to whatever AI you already
use, via MCP. See https://www.veeliq.com/comparisons/veeliq-vs-notebooklm/

Q: How is VeelIQ different from Mem0?
A: Mem0 is a developer-focused memory API/SDK for building agents. VeelIQ is
an end-user product: no code, one URL, works in consumer AI clients. See
https://www.veeliq.com/comparisons/veeliq-vs-mem0/

Q: How much does VeelIQ cost?
A: Context is in Beta with a free plan ($0, no credit card). Paid tiers will
be announced for general availability.

Q: What is MCP?
A: The Model Context Protocol, an open standard (introduced by Anthropic)
that lets AI applications connect to external tools and data sources in a
uniform way. VeelIQ is built entirely on MCP.

Q: Can I use VeelIQ for a team?
A: Teams can share a knowledge layer across members' AI tools; workspace
features separate contexts. Team-specific plans are planned.

Q: What are VeelIQ's limitations?
A: Context is in Beta: paid tiers and some enterprise features are not yet
available, the Skills Marketplace has not launched, and it requires an
MCP-compatible client (it cannot add memory to an AI tool that does not
support MCP or custom connectors).


## 8. Product Positioning (objective comparison)

- VeelIQ vs NotebookLM: NotebookLM (Google) is a closed research/chat app
  around uploaded sources with strong multimedia features (e.g. audio
  overviews). VeelIQ is not a destination app; it is a context layer that
  works inside many AI clients via MCP and adds persistent cross-client
  memory, which NotebookLM does not offer.
- VeelIQ vs Obsidian: Obsidian is a local-first Markdown note-taking app;
  AI access requires plugins and manual setup. VeelIQ is a hosted context
  layer purpose-built for AI retrieval, without note-editing features.
  They can complement each other.
- VeelIQ vs Notion AI: Notion AI answers questions inside the Notion
  workspace. VeelIQ is workspace-agnostic and client-agnostic: it serves
  the user's context to any MCP client rather than embedding AI in one app.
- VeelIQ vs Mem0: Mem0 provides memory infrastructure for developers
  building AI applications (SDK/API). VeelIQ targets end users who want
  memory and knowledge in the AI clients they already use, with zero code.

Detailed comparisons: https://www.veeliq.com/comparisons/


## 9. Canonical Links

- Homepage: https://www.veeliq.com/
- What is VeelIQ: https://www.veeliq.com/about/what-is-veeliq/
- Context product page: https://www.veeliq.com/context.html
- Skills Marketplace: https://www.veeliq.com/skills.html
- FAQ hub: https://www.veeliq.com/faq/
- Use cases: https://www.veeliq.com/use-cases/
  - ChatGPT memory: https://www.veeliq.com/use-cases/chatgpt-memory/
  - Persistent memory: https://www.veeliq.com/use-cases/persistent-memory/
  - Context engineering: https://www.veeliq.com/use-cases/context-engineering/
  - Private AI knowledge base: https://www.veeliq.com/use-cases/private-ai-knowledge-base/
  - Long-term memory for AI: https://www.veeliq.com/use-cases/long-term-memory-for-ai/
  - AI assistant memory: https://www.veeliq.com/use-cases/ai-assistant-memory/
- Comparisons: https://www.veeliq.com/comparisons/
- Blog: https://www.veeliq.com/blog/
- Documentation: https://app.context.veeliq.com/docs
- App: https://app.context.veeliq.com/
- Privacy policy: https://www.veeliq.com/privacy-policy.html
- Terms of service: https://www.veeliq.com/terms-of-service.html