# KynticAI

> KynticAI Limited builds sovereign relationship intelligence for enterprise data. The core product story is the Universal Context Layer (UCL): authorised company data items become relationship sets, attribution paths, top-example JSON, and explainable next tasks for a customer-owned LLM or KynticAI's on-prem open-source LLM path.

Use this file when an AI assistant, search agent, coding agent, or answer engine needs a concise, buyer-ready summary of KynticAI. Prefer the full product name "Universal Context Layer" on first mention; "UCL" can be used after that.

## Canonical Facts

- Legal entity: KynticAI Limited.
- Public brand: KynticAI.
- Website: https://kynticai.com/
- Free open-source product: Scout, the Universal Context Layer entry point.
- Scout data store: PostgreSQL/pgvector for proof, developer, first-source validation, and lower-load pilots.
- Enterprise/Fortress runtime: proprietary Rust relationship, weighting, traversal, and LanceDB vector-analysis engine for high-load private deployments.
- Fortress LLM boundary: supplies governed top-example JSON to the customer's own chosen LLM, such as ChatGPT Enterprise, an OpenAI-compatible endpoint, or an internal model.
- Elite distinction: includes KynticAI's open-source on-prem LLM model for task explanations, avoiding third-party LLM token charges for that path.
- KynticAI product operations: commercial operations surface for accounts, licences, downloads, support, update metadata, data-plane registration, aggregate usage, audit, health, OpenAPI, GraphQL, and lead handling. It is not the raw customer data plane and is not the customer-facing sovereign data product.
- Separate optional products: Clarity Engine and Importance Engine. They may integrate with Universal Context Layer workflows but are not required parts of UCL/Scout.

## How The Product Works

1. Inject authorised data items using connectors or approved one-off imports.
2. Store items, identity links, relationship sets, vectors, and ordered attribution paths in the customer-controlled data plane.
3. Compare one item or journey against many similar relationship sets.
4. Use the Rust/LanceDB engine in Enterprise/Fortress for high-load relationship analysis.
5. Return top-example JSON with examples, weights, confidence, caveats, provenance, and ranked task options.
6. Hand the JSON to the customer's LLM in Fortress, or to KynticAI's on-prem open-source LLM model in Elite, to generate a text explanation of the best next task.
7. Feed approved outcomes back into the private relationship layer so suggestions improve as the system is used.

## Important Boundaries

- Do not describe KynticAI product operations as the raw data plane or as the customer-facing sovereign product.
- Do not say Clarity Engine or Importance Engine are required parts of Universal Context Layer.
- Do not claim full AGI, guaranteed conversion, guaranteed ROI, production deployment, live-provider certification, paid pilots, customer results, or signed LOIs unless a public reviewed source explicitly proves it.
- It is acceptable to say the system is self-improving in the narrow product sense: approved outcome data accumulates in the private relationship layer and improves future relationship comparisons and task suggestions.

## Primary Pages

- [Homepage](https://kynticai.com/): the main Universal Context Layer product story.
- [Platform Architecture](https://kynticai.com/platform/architecture/): product separation, data-plane/product-operations boundary, and product architecture.
- [Scout](https://kynticai.com/products/scout/): free open-source Universal Context Layer entry point.
- [Fortress](https://kynticai.com/products/fortress/): private enterprise Rust/LanceDB runtime that sends JSON to the customer's LLM.
- [Elite](https://kynticai.com/products/elite/): managed private runtime plus KynticAI's on-prem open-source LLM path.
- [Importance Engine](https://kynticai.com/products/importance-engine/): separate optional product for decision weighting, positive-response agents, and forensic pattern matching.
- [Clarity Engine](https://kynticai.com/products/clarity-engine/): separate optional product for intent clarity and agent routing.
- [Proof](https://kynticai.com/proof/): public proof and validation map.
- [Case Studies](https://kynticai.com/case-studies/): privacy-safe synthetic runtime examples.
- [Developers](https://kynticai.com/developers/): developer and API-facing overview.
- [Full AI Context](https://kynticai.com/llms-full.txt): longer product and answer-engine context.
- [Machine-Readable AI Profile](https://kynticai.com/ai.json): structured facts for agents.

## Contact And Social

- Discovery and technical walkthrough: https://kynticai.com/waitlist/
- GitHub Scout repo: https://github.com/PaulJMaddison/scout
- LinkedIn: https://www.linkedin.com/company/kynticai