Order from
complexity.

An operating layer that connects your people, data, technology, governance and workflows. Aligned to your strategy and business value, and kept current as the technology and the work evolve.

Built for businesses and teams  ·  whose work is judgement-led, and where the handoffs between people and AI need to be deliberately designed and curated.

Built around your people, not the technology.
Scroll

One operating layer.

One place for your leadership, your AI leads and your teams to see and run how the work gets done with AI. It sits on your tools and platforms, aligns key pieces into a working structure, and helps keep it current.

MeridianAI Operating HubIllustrative environmentKALEI
HomeStrategyAgentsWorkflowsAI ToolsOpportunitiesKnowledgeGetting StartedGovernanceAdoption & ROI
Meridian × Kalei
The Meridian AI Operating Hub
Strategy & Priorities
Value, priorities and design principles, agreed upfront.
Agents Library
14 sanctioned agents. 5 live, 9 in build.
Workflows
Craft amplified by AI. People keep judgement.
Adoption & ROI
Time saved and return, against held baselines.
IntelligenceFrom Kalei
Agents can now run inside your own security boundary. Anthropic’s managed agents support self-hosted sandboxes (public beta) and private MCP connections this month. Removes the main blocker to pointing Meridian’s research agents at confidential client material. Tech
AI-use disclosure is entering RFPs. Seen across Kalei engagements this quarter: procurement teams increasingly ask for a documented AI policy and a named human-review chain. Firms that can show governance, not just tooling, are starting to win on it. Market move
RecommendationsFrom Kalei
Pilot the self-hosted sandbox for External Research Intelligence. It currently works from public sources only; inside the firm’s boundary it can draw on client briefs too. Two-week pilot, IT sign-off first. Optimise
Turn governance into a selling point. Package the review chain, data-handling rules and agent register from the Governance section into a client-facing annex for tenders. Low effort — the material already exists in the hub. Strategic
Your hub, with Kalei’s intelligence and recommended actions refreshed on the cadence you set.
01

Strategy & priorities

Agreed key value drivers, your business priorities, and the design principles that every later decision is measured against.

02

Agents

Every sanctioned agent in one library: creates consistency and a mechanism to validate and improve as a team.

03

Workflows

Your highest-value processes, rebuilt end to end. Connecting agents, shared context, knowledge bases and designed human-AI handoffs.

04

AI tools

The approved toolkit for the firm, with what each tool is for, what it costs, and what the rules are. Kalei helps scan the market and recommends new ones.

05

Opportunities

A tiered roadmap of the new market moves that AI opens up, scored on value and feasibility.

06

Knowledge

Proprietary knowledge bases, best practice and context data, embedded into workflows and agents or used on their own, each with an assigned owner.

07

Governance

Ownership and decision rights, plus the policies, data handling and privacy instructions that travel with the work.

08

Adoption & ROI

Usage, time saved and return, all measured against the baselines set at the start.

The platform is the destination.
A structured onboarding gets you there.

We help connect and activate the right pieces in the right order, and the platform goes live only when it is ready for your team.

Step 01
01

Discover, align and prioritise

We begin with the value at stake rather than the tools. We agree your priorities, set the design principles that guide every later choice, and map how the work really happens now. Then we score where AI earns its place, so effort lands where it moves the business and is held back where judgement should stay human.

Value at stakePriorities & principlesWhere AI earns its place
Step 02
02
Efficiency play

Redesign workflows and assemble agents

We redesign your priority workflows with the handoffs between people and AI, then assemble the agents to run them, or leverage those you already have. Technology choice is based on the stack you already own. This is where the time and margin come back.

Workflows redesignedAgents builtMargin recovered
Step 03
03
Growth play

Find new market opportunities

Efficiency is only part of value. Through market analysis and innovation workshops, we identify the new products and services that AI now makes possible for you, and set them on a tiered roadmap scored on value and feasibility. The same capability that saves time also opens new ground.

Market analysisInnovation workshopsOpportunity roadmap
Step 04
04

Govern, then go live

Every workflow, agent and knowledge base gets a named owner and clear decision rights. Policies, data handling and privacy instructions travel with the work rather than sitting in a separate document. Client material stays inside your environment. The instance goes live once it is ready.

Named ownershipPolicy with the workPrivacy first
Step 05
05

Keep it current

Technology moves faster than most companies can. Kalei closes that gap. We track the vendor and model landscape for you, watch market dynamics and token costs, and send recommended actions tailored to your stack, re-aligning prompts, models, data and human-AI handoffs on a cadence you set, as both the technology and your people move on.

Tech and market scanRecommended actionsSupport on your cadence

What breaks when AI goes in without it.

None of these are technology problems. They are structural, and they compound.

01 — Sprawl

AI sprawl

Tools and prompts scatter across individuals. No shared truth, and duplicated effort across teams.

Kalei · One sanctioned library
02 — Ownership

Ownership gaps

Workflows live in nobody's job description. No accountability for maintaining the quality and managing risks.

Kalei · Named owners on everything
03 — Process

One-size-fits-all

Low-stakes work is over-controlled. High-stakes work flies without guardrails. Friction in the wrong places.

Kalei · Controls matched to stakes
04 — Handoffs

Unclear handoffs

AI recommends, people rubber-stamp. Decision rights drift without anyone deciding, and accountability goes missing.

Kalei · Handoffs designed in
05 — Outcomes

Activity over outcome

Adoption rates look good on a slide. They say nothing about whether the business actually moved.

Kalei · Measured against baselines
06 — Currency

Stuck in time

The build that fit last year's models quietly ages. Without a layer that keeps it current, today's investment becomes tomorrow's legacy.

Kalei · Kept current for you

Built around your people, not the technology.

Most AI programmes start with the tools. We start with the work, and with how your people actually think, decide and collaborate. Drawing on a background in digital anthropology and the human side of AI, we design for judgement, trust and adoption: what to hand to AI, what stays human, how quality is held, and how people come to reach for the system rather than work around it. AI works as an instrument of your people’s judgement, never a replacement for it. Adoption is a human problem first.

See what your operating layer could look like.

Many firms run AI in pockets, a tool here and a prompt there, with no shared picture. Tell us where you are today, and we will show you a tailored view of how your operating layer could work for your business or your team, and what it would take to build it. You leave with a working picture of the destination and the path to it.

Prefer to talk first? Request an introduction and we will take it from there.