8 by 8
Trustworthy AI · for the people doing good

Context makes AI trustworthy.

8 by 8 helps charities, not-for-profits, small businesses and emergency services build the right context that makes their AI trustworthy and reliable — without an AI team of your own.

Tell us your challenge See how it works
The problem

The gap between hype and AI you can rely on

Before AI even comes up, the same everyday things get in the way of doing good:

Everything by hand

Tedious manual tasks and spreadsheet after spreadsheet — the same work, repeated week after week.

Data all over the place

Your information sits in different systems that don't talk to each other, so nothing is joined up.

Always firefighting

Hours lost fixing what breaks — time pulled away from the core work of doing good.

AI
Your work is scattered across spreadsheets, documents, slide decks and data warehouses. We bring that context together — so AI can actually rely on it.

AI is meant to lift this load. But raw AI is only as good as the context it's given — and the organisations that could do the most good often have the fewest resources to build it. Without that context, AI:

  • Guesses instead of grounding — confident answers that are wrong.
  • Puts trust at risk — with the people you serve and your donors.
  • Demands expertise you don't have — usually a whole AI team.
  • Stays out of reach — leaving the best causes locked out.
Case studies

See what grounded AI makes possible

Illustrative, interactive demonstrations — built to show the shape of the work and how trust is engineered in.

Speed to value

Parallel AI & data transformation

Build AI alongside the data platform, not after it — so the spend pays for itself sooner (~month 7, not ~month 26), with the cost of waiting shown as an irrecoverable gap.
Emergency services

Spatial intelligence for emergency services

Calls, vehicles and incidents on one live operational picture — connected as a knowledge graph, placed in time and space, and replayable second by second.
Investment decisioning

ML+AI-empowered driver trees

Decompose value into drivers, with AI and ML estimating impact, Shapley attribution and Monte Carlo search answering “which investment, where, gives the best ROI?”
Search & rescue

Overlaying cell tower data and LiDAR for search & rescue

Read the land and the last faint trace of a phone’s signal together, to point searchers at the likeliest ground first. A research proposal.
Supply & demand management

From spreadsheets to a plain-language agent

Hundreds of linked, hand-keyed spreadsheets — the process pulled out, confirmed by the client's own experts, and turned into one knowledge base behind an agent you simply talk to. 1,600% ROI; the work of a team of five.
Your challenge

Have a challenge in mind?

These are illustrative. If you have a problem where grounded AI could help, tell us — it could be the next case study.
What we do

Software plus hands-on guidance

We build, manage and deliver the right context to your AI — and help your team every step, so you don't need to become AI experts.

Build

From what you have

Turn your policies, processes, conversations, documentation and data into structured, reusable context.

Manage

Kept current

Context stays owned, up to date and under control as your organisation evolves.

Deliver

Grounded answers

The right context reaches each AI request, so answers stay grounded in what matters.

Check

Stays reliable

We monitor that answers remain accurate and dependable over time.

How it works

From your world to AI you can trust

Check · monitor & improve Your worldpolicies · data · docs Buildextract & codify Deploylocal or hosted Managegovern & update Deliverground each answer Trustworthy AIreliable · explainable

Hover or tap a step.

How we work

Governance and design, first.

We design for trust from the outset — governance, grounding and accountability are built in: never bolted on, never an afterthought. The rest of our method makes that real:

01

Built on research

We apply the latest evidence on what makes AI reliable and accurate.

02

Informed by others

We consult and learn from researchers, practitioners and the organisations we serve.

03

Made practical

We turn learnings into tools and guidance a small team can readily use.

04

Start small, learn fast

We begin focused, prove value, and scale only what works.

Who we serve

Made for the people doing good

Charities

Help more people without risking AI getting it wrong with the communities you serve.

Not-for-profits

Do more with the resources you have — reliable AI that fits a lean team.

Small businesses

Compete and grow with AI's leverage, without becoming an AI expert.

Emergency services

When lives depend on it, AI must be reliable — grounded plans you can act on under pressure.

What you get

Reliable AI, within reach

Trust without the team

We handle the context and reliability work, so you don't have to become AI experts.

Do more good

Serve more people without adding headcount.

Protect the people you serve

Accurate, explainable answers you can stand behind.

Made for your budget

The rigour big organisations pay a fortune for — made affordable.

Grounded in research

Backed by evidence, so the reliability we promise is real.

No lock-in

Model-agnostic by design — your context and data remain yours.

Why you can trust it

Built to be relied on

Research-led

Institutions Publications Researchers Industrial experiment Methods Your AI

Grounded, not guessed

fail → retry Question Context AI agent Response Validator Grounded ✓

Your data stays yours

REGION YOUR ENVIRONMENT Your data Local AI Answer

Built for small teams

You 8 by 8 Just use it

Our story

Why “8 by 8”

The board

A chessboard is 8×8 — 64 squares. In chess the whole board is visible, so every move is deliberate and can be explained. AI without context plays blind. 8 by 8 gives it the full picture, so every answer is a considered move.

The 64 hours

Friday evening to Monday morning is exactly 64 hours — the weekend. The quiet stretch when the steady, unglamorous work gets done. That's 8 by 8: we do that work in the background, so your small team doesn't have to.

Trust is earned through steady, careful work. Reliability isn't loud.

Community

We give back

Doing good isn't a side project for us — it's the point. We build in the open and give our time to the causes that need it.

We donate code

We open-source tools and components, so the whole community — not just our clients — can build AI it can trust. A couple of ours are below.

Pro-bono work

Small asks? We simply help — free. And we dedicate capacity to pro-bono projects for charities and good causes that couldn't otherwise reach trustworthy AI.

We reinvest

Paid work — including from larger institutions — helps fund the pro-bono help and the research that keep good causes moving forward.

Free AI learning sessions — every weekend

Open, hands-on sessions to help anyone get started with AI they can trust — free, every weekend, no jargon. Bring your questions.

Join a session

A couple of the tools we've open-sourced so far — free for anyone to use, MIT licensed. More for the technically minded:

Tell us your challenge — big or small.

If it's small enough, we'll simply help, pro-bono. Otherwise, we'll show you what's possible — and how to make your AI something you can rely on.

Chat now