AI that fits around your business

Your knowledge.
Your permissions.
At your fingertips.

Enterprise search that indexes your documents in place, respects your existing permissions, and answers questions with sourced citations. Nothing leaves your network.

01

Search results with ranked documents

The Problem

Five problems. One answer.

Your team loses hours every week to tasks that one knowledge system can handle in seconds.

Searching scattered filesAnswering repeated questionsOnboarding new startersChasing colleagues for answersNavigating policies & complianceAthenaOne AI. Every answer.

1.8 hours a day digging through drives, inboxes, and chat threads for one answer.

HR, IT, and Legal field the same queries week after week.

Weeks of ramp-up reading tribal knowledge that lives in people’s heads.

20% of the week tracking down the one person who knows — and they’re in a meeting.

The right version of the right policy — buried three folders deep.

How we built it

Nothing moves. Nothing changes. Everything improves.

Fits into your existing infrastructure. No new data stores, no new permission systems.

No data warehouse needed

Indexes your content in place. Files stay on the same shares, same folders, same policies.

Your security model, as is

Integrates with Active Directory or your existing SSO. No parallel permission system to manage.

PII never enters the index

Content is anonymised during indexing. Only the meaning and structure needed to find answers.

Features

Everything your team needs

Every answer links back to the original document

Combines keyword and meaning-based search to find the most relevant passages, then generates a clear answer with citations.

  • Every answer cites the source document and page
  • Keyword and semantic search combined for precision
  • Results ranked by relevance, not recency
  • Reduces hallucination by grounding responses in your data

How it works

From raw documents to sourced answers in minutes

The entire pipeline runs on your hardware.

01

We deploy on your hardware

Our engineers install Athena on your infrastructure. Docker containers, GPU detection, and networking configured in days.

02

Connect your sources

Point Athena at SharePoint, shared drives, email archives, code repos, or databases. 30+ connectors available.

03

Automatic indexing begins

Documents are chunked, embedded, and indexed in the background. The search index stays current as content changes.

04

Your team starts asking questions

Ask in plain English. Get a sourced answer with links to the original documents, in seconds.

Capabilities

What makes Athena different

Where other AI tools fall short, and how Athena does it differently.

Runs entirely on your infrastructure

Every query is processed on hardware you control. No data leaves your network, ever. Air-gap compatible for regulated environments.

Open-source models, no vendor lock-in

Built on Qwen, Mistral, DeepSeek, and Llama. No proprietary model dependency. Swap, upgrade, or fine-tune at any time.

Optimised for your hardware automatically

Detects your GPU configuration and selects the best model, context length, and batch size. No manual tuning required.

Full audit trail

Every query, every access event, and every answer is logged. GDPR and HIPAA-ready compliance reporting out of the box.

Flat-fee pricing, no per-query charges

One monthly licence covers unlimited users and unlimited queries. Costs stay predictable as adoption grows.

Production-grade from day one

Built-in monitoring, alerting, TLS termination, and health checks. 16 containerised services designed for reliability.

Comparison

Athena vs. the alternatives

Cloud RAG APIs require sending your data out. DIY takes months of engineering. Athena gives you both without the trade-offs.

FeatureAthenaCloud RAG APIDIY On-Prem
No data warehouse or migration needed
Uses existing permissions as isManual
PII anonymised in index
Data stays on your infrastructure
No data sent to third parties
No usage-based charges
Automatically optimised for your hardware
Combined keyword and meaning-based searchVaries
Relevance-based result rankingPaid add-on
Audit logsPaid tierManual
Air-gap compatible
Production-ready on installation

Under the hood

Open-source models, closed network

Every component runs on your infrastructure. Open-source models. None of it calls home.

AI & Language Models

  • vLLM

    Runs AI models locally. No cloud dependency

  • Qwen · Mistral · DeepSeek · Llama

    Leading open-weight model families, running locally

  • Sentence Transformers

    Local document processing. Nothing leaves your network

  • Vision Models

    Image, diagram, and scanned document understanding

Search & Data

  • ChromaDB

    Sub-100ms search results

  • PostgreSQL

    Users, sessions, metadata, and full audit trail

  • Redis

    Query caching and inter-service messaging

  • Hybrid search

    Keyword and meaning-based search, ranked by relevance

Infrastructure

  • Docker

    16 independent services, consistent across environments

  • FastAPI

    Python API with auto-generated reference docs

  • Nginx

    Reverse proxy with TLS 1.3, HSTS, and security headers

  • Prometheus + Grafana

    Built-in monitoring, dashboards, and alerting

Connectors

  • Microsoft 365

    SharePoint, OneDrive, Teams, Azure DevOps

  • Atlassian · Notion · Slack

    Confluence, Jira, and collaboration tools

  • Git providers

    GitHub, GitLab, Bitbucket, with commit history

  • Cloud storage + databases

    S3, Google Drive, Dropbox, Box, PostgreSQL, MySQL

Security & Compliance

  • Azure AD / SSO

    Your existing identity provider, as is

  • AES-256-GCM encryption

    Every document encrypted at rest

  • GDPR & HIPAA audit logs

    Every query and access event recorded

  • Zero external calls

    No telemetry, no callbacks. Everything stays on your network

Athena ROI Calculator

Calculate your savings with Athena

See how much your organisation could save by cutting the time staff spend searching for information.

200
1050,000
£45,000
£20,000£150,000
5 hrs
1 hr20 hrs

Net annual savings with Athena

£995,250

per year (after Athena licence cost)

47,667

hours recovered per year

£2,000

licence / retainer per month

£12,000

one-time setup

£1,125,000 currently wasted per year

Talk to us about Athena pricing

Plans

Simple, predictable pricing

Flat monthly licence plus one-time setup. No per-seat or per-query charges.

Limited Availability

Early Adopter Programme

90% off

Lock in 90% off implementation, guaranteed for three years. No per-seat fees. No per-query charges. The same flat rate as you grow.

Apply for early access

Standard

~£2,000/mo

Full platform licence with support, updates, and SLA.

  • Athena platform licence
  • All platform connectors (15+)
  • Advanced search with intelligent ranking
  • Multi-user access controls & SSO
  • Audit logging
  • Security patches & updates
  • ½ support day / month
  • 3-business-day SLA
Book a demo
Most Popular

Enterprise

~£5,000/mo

Priority SLA, dedicated support, ISO 27001 and SOC 2 ready.

  • Everything in Standard
  • Offline / air-gapped deployment
  • On-site installation & training
  • 2 support days / month
  • 1-business-day SLA
  • Dedicated account manager
  • Workshops & advanced training
  • Compliance (ISO 27001, SOC 2)
Contact sales

One-time implementation fee

Based on organisation size. Includes deployment, integration, training, and 90-day support.

SizeUsersStandardEarly Adopter
SmallUp to 500 usersFrom £5,000From £500
Medium500 – 2,500 usersFrom £25,000From £2,500
Large2,500+ usersFrom £75,000From £7,500
Full deployment & configuration
SSO integration
Document source setup
Security review
User training
90-day post-launch support

FAQ

Frequently asked questions

Can't find what you're looking for? Email us at hello@athenaai.uk

Ready to own your AI?

30-minute demo.
Your data.
Your hardware.

Have feedback on Athena? Feedback