# Listening Post

> Your company's private AI. Listening Post is the complete platform for running AI on your terms — models, tooling, agents, and infrastructure that keep your data inside your perimeter and out of someone else's training set. SOC 2 audited, patented orchestration, deployable on-prem, in private cloud, or air-gapped.

Listening Post lets enterprises run AI without giving up data control. It deploys as a sidecar alongside existing systems (no rip-and-replace), reads data in place over mTLS with scoped IAM, and blocks egress of prompts, embeddings, and keys to public LLMs. Customers control their own keys, audit logs, and egress policy.

## Pages

- [Home](https://listeningpost.ai/index.html): Positioning, the data-control problem, the four-pillar solution, how the sidecar architecture works, and security/compliance overview.
- [Product](https://listeningpost.ai/product.html): The platform's four pillars — Models, Tooling, Agents, Infrastructure — and how they collapse a six-vendor private-AI stack into one.
- [Security](https://listeningpost.ai/security.html): Security model and compliance posture — SOC 2 Type II, customer-managed keys, audit-ready logging, request flow, and on-prem / private cloud / air-gapped deployment modes.
- [Data](https://listeningpost.ai/data.html): The Listening Post Graph — a behavioral data layer covering 250M+ verified people and 60M+ businesses, 81,000+ signals per person, queryable in plain English from agents.
- [Request a Demo](https://listeningpost.ai/demo.html): Scope a private pilot — a 30-minute working session mapping your perimeter, priority use case, and a 60-day path to production.

## Platform pillars

- Models: Open-weight (Mistral, Gemma, GPT-OSS) and frontier models deployed in your environment. No prompt egress, no fine-tuning leakage.
- Tooling: Evaluation, observability, retrieval, and guardrails as native primitives.
- Agents: Long-running workflows that read your systems in place and never copy data out. Audit-ready by default.
- Infrastructure: Kubernetes-native inference, serving, and orchestration. On-prem, private cloud, or air-gapped. Bring your own KMS/HSM.

## Compliance

- SOC 2 Type II (current)
- HIPAA (configurable)
- FedRAMP Moderate (roadmap)
- NIST AI RMF aligned (current)
- Patented orchestration architecture (issued)

## Contact

- Request a demo: https://listeningpost.ai/demo.html
- Email: info@listeningpost.ai