Skip to content
Custom · configure under the three minds

Custom agents, under the three minds.

Simon, Maeve and Hugo are the three minds — the only three. Under each, configure as many agents as your business needs, each its own employee. When a role none of the templates cover comes up, build a fully custom agent from scratch — always as a Simon (web), a Maeve (phone) or a Hugo (internal). Give it a persona, skills, a knowledge scope and an access policy. It runs on the same shared Brain, and it’s measurable just like the rest of the team.

How it works

Four decisions and your agent is on the team.

You don’t write prompts or wire pipelines. You pick which mind it is, make four plain-language decisions, and NeoMind assembles a working, measurable agent on your Brain.

Persona

Give it a name, a role and a tone of voice — how it should sound and behave when it talks to people.

Skills

Pick what it can do — answer questions, capture leads, book, route, summarise — from the same skills the standard three use.

Knowledge scope

Choose which parts of your Brain it can read — so it’s an expert on exactly what it needs, and nothing it shouldn’t see.

Access policy

Set where it operates and what actions it’s allowed to take — you stay in control of the boundaries.

The result: a working agent on your shared Brain — consistent with the rest of the team, updated the instant your knowledge changes, and scored on every conversation.
Ideas to start from

A few roles teams configure.

Illustrative starting points — you can shape an agent around almost any repeatable, knowledge-driven role, under whichever mind fits.

Customer

Onboarding guide

Walks new customers through setup, answers product questions from your docs, and hands off to a human when something needs configuring.

Internal

IT help desk

Answers common IT and tooling questions for staff, points to the right runbook, and raises a ticket when it’s beyond self-serve.

Operations

Bookings & scheduling

Handles a specific service line’s enquiries and bookings end-to-end, scoped to just that part of your knowledge.

Same guarantees

Custom, but never a loose cannon.

A custom agent is built the same way as a sales Simon or an HR Hugo — it’s still one of the three minds, just configured from scratch for a role of your own. It inherits everything that makes the default agents trustworthy: it answers only from the knowledge you scope it to, with citations, and it’s measurable — its own objective and weighted KPIs, scored on every conversation by an AI judge and rolled into a scorecard. See how measurement works →

And it respects the same bright line: it does the routine work, but judgement calls and anything binding stay with your team. NeoMind never signs, commits, or makes a binding promise on your behalf. Every agent draws from the same shared Brain, so the whole team stays in sync. How the Brain works →

Frequently asked

Custom agents, answered.

It’s an agent you configure under one of the three minds — a Simon for your website, a Maeve for your phone, or a Hugo for your team. You give it a persona, a set of skills (answer, capture, book, route, summarise), a knowledge scope (which parts of the Brain it can read) and an access policy (where it operates and what it can do). It runs on the same shared Brain as every other agent.

It isn’t a different kind of thing. Simon, Maeve and Hugo are the three minds — the only three — and every agent, default or custom, is one of them. A custom agent is simply one you’ve configured from scratch for a role of your own (a returns Simon, an onboarding Hugo) rather than starting from a template. It uses the same building blocks, the same shared Brain, and the same scoring.

Yes. Every agent draws from the same shared Brain, so it stays consistent with the rest of your team and updates the moment your knowledge changes — no separate training.

Yes. Each agent has a knowledge scope and an access policy, so you decide which information it can read and which actions it can take. Sensitive areas stay out of scope unless you grant them.

Yes. A custom agent carries its own objective and weighted KPIs, scored on every conversation by an AI judge and rolled into a scorecard — exactly like the default Simon, Hugo and Maeve.

No. Like every NeoMind agent, a custom one does the routine work but keeps judgement and anything binding with your team. NeoMind never signs, commits, or makes a binding promise on your behalf.

Prefer to start with the ready-made roster? Meet Simon, Hugo and Maeve — or see what to look for in the best AI agents for SMBs in 2026.

Get started

Configure the agents your business needs.

Pick a mind, define a role, scope its knowledge, and it’s on the team — running on your shared Brain, measurable from day one.