# InferTheta

> InferTheta is a decision infrastructure platform that uses probabilistic graphs and Bayesian inference to govern AI agents and autonomous hardware.

InferTheta lets engineers design, train, and deploy Bayesian / probabilistic graph models visually (no code), then run exact inference for AI agents, autonomous hardware, and human decision support.

Prefer the Markdown mirrors below when extracting product facts; they are complete without executing JavaScript.

## Agent API (start here if you are an AI agent)

Use the production HTTP Agent API for tools, inference, and credits. This website is human marketing + modelling docs.

- Base URL: https://infertheta-agent-engine-production.up.railway.app/
- Agent discovery: https://infertheta-agent-engine-production.up.railway.app/llms.txt
- OpenAPI: https://infertheta-agent-engine-production.up.railway.app/openapi.json
- Pricing: https://infertheta-agent-engine-production.up.railway.app/v1/pricing
- Policy: https://infertheta-agent-engine-production.up.railway.app/v1/policy

### Access (no Auth0)

1. Free sandbox: `POST https://infertheta-agent-engine-production.up.railway.app/v1/sandbox` → `apiKey` + small credits
2. Pay-per-call: `GET https://infertheta-agent-engine-production.up.railway.app/v1/pricing` → `POST https://infertheta-agent-engine-production.up.railway.app/v1/checkout` `{ "pack": "starter" }` → human opens `checkoutUrl` → `GET https://infertheta-agent-engine-production.up.railway.app/v1/checkout/{sessionId}` returns `apiKey` once
3. Call with `Authorization: Bearer <apiKey>` or `X-API-Key: <apiKey>`
4. Top up: `POST /v1/checkout` `{ "pack": "builder", "topUp": true }` with existing key (human pays)

Agents must not complete card checkout or SMS verification themselves — hand `checkoutUrl` to a human.

### Typical tool workflow

1. Obtain `apiKey` (sandbox or checkout)
2. `POST /v1/networks` — create or upload a Bayesian network
3. `POST /v1/networks/{id}/build` or `/nodes` `/edges` `/cpt` — design + fill CPTs
4. `POST /v1/networks/{id}/evidence` — observe facts
5. `POST /v1/networks/{id}/inference` — compute posteriors
6. `POST /v1/networks/{id}/evaluate-decision` — choose action
7. `POST /v1/networks/{id}/explain` — auditable explanation for a human

## Primary pages (humans + product facts)

- [Home](https://infertheta.com/): Product overview (SPA). Prefer the Markdown mirror for agents.
- [For agents (human explainer)](https://infertheta.com/for-agents): Why fund InferTheta for your agents; links to machine entry
- [Home (Markdown mirror)](https://infertheta.com/llms-home.md): Full landing-page copy, thesis, offerings, sectors, advantages, and FAQ
- [Documentation / User guide](https://infertheta.com/documentation/): Full plain-language software guide (static HTML)
- [Full user guide (Markdown)](https://infertheta.com/llms-full.md): Same guide as clean text for AI agents
- [Case studies](https://infertheta.com/case-studies): Sector applications
- [Contact / demo](https://infertheta.com/#contact-form): Request access or engagement

## What InferTheta is / is not

- Is: decision infrastructure using probabilistic graphs and Bayesian inference to govern AI agents and autonomous hardware
- Is not: a chat LLM, a black-box neural policy, or a drop-in replacement for your existing agent/robot stack
- Typical pattern: observe → infer on the graph → act within explicit decision bounds

## FAQ

- **What is InferTheta?** InferTheta is a decision infrastructure platform that uses probabilistic graphs and Bayesian inference to govern AI agents and autonomous hardware.
- **How is InferTheta different from an LLM agent?** LLM agents generate fluent text but can hallucinate and are hard to audit. InferTheta sits as a decision layer: observations update a probabilistic graph, and actions are chosen from exact posteriors and explicit decision rules you can inspect.
- **Who is InferTheta for?** Teams building AI agents or autonomous hardware in high-stakes domains—defense, cybersecurity, finance, robotics, manufacturing, supply chain, and critical infrastructure—that need governed, explainable decisions under uncertainty.
- **Do I need to write code to use InferTheta?** Humans can design graphs visually in the modelling engine. Agents should use the production HTTP Agent API (sandbox key or paid credits)—no Auth0. You can also export standardized JSON or idiomatic Python for edge / embedded runtimes.
- **How do AI agents call InferTheta?** Start at https://infertheta-agent-engine-production.up.railway.app/llms.txt or https://infertheta-agent-engine-production.up.railway.app/openapi.json. POST /v1/sandbox for a free API key, or POST /v1/checkout for paid credits (a human completes Stripe Checkout). Then create networks, set evidence, run inference, and request explanations with Authorization: Bearer <apiKey>.
- **Does InferTheta replace my existing agents or robots?** No. Agents and machines call InferTheta graphs as a reasoning / governance tool: observe → infer → act, with bounded and auditable outputs.

## Featured case studies

- [Counter-UAS guardrails for AI agents](https://infertheta.com/case-studies/defense-counter-uas)
- [Bayesian sensor fusion in contested environments](https://infertheta.com/case-studies/defense-contested-environments)
- [Cyber threat prioritization](https://infertheta.com/case-studies/cybersecurity-threat-prioritization)
- [Probabilistic supply chain disruption simulation](https://infertheta.com/case-studies/supply-chain)
- [PE fund secondaries ZOPA pricing](https://infertheta.com/case-studies/investment-pe-secondaries)
- [Thermal runaway early warning](https://infertheta.com/case-studies/engineering-thermal-runaway)

## Documentation sections

- [1. What Bayesian networks / Probabilistic graphs are (simply put)](https://infertheta.com/documentation/#1-what-bayesian-networks-probabilistic-graphs-are-simply-put)
- [Nodes, edges, and uncertainty](https://infertheta.com/documentation/#nodes-edges-and-uncertainty)
- [Why the diagram saves so much work](https://infertheta.com/documentation/#why-the-diagram-saves-so-much-work)
- [Two ways to read the same diagram](https://infertheta.com/documentation/#two-ways-to-read-the-same-diagram)
- [What you do with a finished network](https://infertheta.com/documentation/#what-you-do-with-a-finished-network)
- [2. The big picture in one minute](https://infertheta.com/documentation/#2-the-big-picture-in-one-minute)
- [3. Getting started](https://infertheta.com/documentation/#3-getting-started)
- [4. Build a model](https://infertheta.com/documentation/#4-build-a-model)
- [4.1 Types of nodes you can add](https://infertheta.com/documentation/#4-1-types-of-nodes-you-can-add)
- [4.2 Connecting nodes (edges)](https://infertheta.com/documentation/#4-2-connecting-nodes-edges)
- [4.3 Editing what a node means](https://infertheta.com/documentation/#4-3-editing-what-a-node-means)
- [5. Ask questions: evidence, scenarios, and inference](https://infertheta.com/documentation/#5-ask-questions-evidence-scenarios-and-inference)
- [6. Learn from data](https://infertheta.com/documentation/#6-learn-from-data)
- [7. Decisions, utilities, and recommended actions](https://infertheta.com/documentation/#7-decisions-utilities-and-recommended-actions)
- [8. Save, share, export, and governance](https://infertheta.com/documentation/#8-save-share-export-and-governance)
- [9. Appearance (UI themes)](https://infertheta.com/documentation/#9-appearance-ui-themes)
- [10. A simple first workflow](https://infertheta.com/documentation/#10-a-simple-first-workflow)
- [11. Glossary (plain language)](https://infertheta.com/documentation/#11-glossary-plain-language)
- [12. Closing note](https://infertheta.com/documentation/#12-closing-note)