AI roadmap

The parts of Serene AI — Serene Core, the Serene assistant, support help, AI Marketing and AI Employees — what is built, what is coming soon, what is planned, and the principles they follow.

Requires setupReferenceFor: Owners, Admins, Managers, Employees, Developers, EvaluatorsLast updated: Oct 18, 2026Introduced: Oct 3, 2026

Built. It answers in early-access workspaces once HyoNova connects its AI provider; until then Serene says it is not available, and nothing is simulated.

HyoNova's AI is one layer — Serene — made of parts that build on each other. This page lists those parts, their status and the rules they all follow.

The parts#

Part	What it does	Status
Serene Core	The shared foundation: acting within each person's permissions, permission-first search with sources, a fixed set of governed tools, approvals, AI disclosure, records of every AI run, and metering in Serene credits	Requires setup
Serene assistant	Answer questions about your company with sources, summarize records with Ask Serene, talk by voice with Talk, create tasks and drafts, and propose changes for your approval. See Using Serene	Requires setup
Meeting summaries	Summaries, decisions and action items after a meeting. Serene can read a meeting's details today; summaries need meeting transcripts, which are planned. See Recording and transcripts	Planned
Support help	Ask Serene on a support conversation summarizes it and suggests a reply; Serene can write an internal note or a reply draft, and sends a reply only with your approval (requires setup). Suggestions shown automatically inside the inbox are planned	Requires setup / Planned
AI Marketing	Help writing campaign and email content, suggesting audiences and designing emails in Hyo Mail. The AI designer for emails already works today	Planned
AI Employees	AI workers hired for a defined job, with a desk, permissions and a budget — starting with the AI Support Employee. See AI Employees	Coming soon
More AI Employee roles	Roles beyond support	Planned

Principles every part follows#

Permissions first. Serene sees and acts only within the permissions of the person using it — or, for an AI Employee, the permissions and knowledge the company granted it. It is never a way around roles.
People approve what matters. Serene may draft and propose; anything that changes important records or sends a message outside the company waits for a person, unless an owner or admin explicitly authorised an AI Employee for low-risk replies. Payments, refunds, permissions, security settings and deleting data are never done by AI.
Your company's data stays yours. Serene works within your company's isolated data, like every other part of HyoNova.
Visible and accountable. AI messages are labelled as AI, customers always see that they are talking to an AI, and every AI run and decision is recorded.
Optional. Owners and admins can turn Serene off for the company, choose what it searches and cap its monthly budget.

See Privacy and safety for how these principles are enforced.

Status#

Feature	Status	Notes
Serene AI	Requires setup	Ask about your company in plain language and get answers with sources, within your own permissions, in a private conversation. Built. It answers in early-access workspaces once HyoNova connects its AI provider; until then Serene says it is not available, and nothing is simulated.
AI Employees	Coming soon	AI workers with a name, a role, a desk in your office, permissions and a budget, working with your team under your rules. Never a user and never a seat. Built and being verified. HyoNova opens hiring only after verifying it in production with a real AI provider.
AI Support Employee	Coming soon	Answers customers in HyoNova Support from your approved knowledge, around the clock, and hands anything sensitive to your team.
AI meeting summaries	Planned	Summaries and action items written by Serene after a meeting.

For everything else on the roadmap, see Coming to HyoNova.

The parts#

Principles every part follows#

Status#

Related