Film-style poster for the Mentornity AI for Startup Programs presentation

Mentornity AI forStartup Programs Mentornity AI for Startup Programs

How incubator and accelerator teams run a cohort with Mentornity AI.

English

What this presentation covers

What is Mentornity AI

An assistant that runs your program

It handles matching, follow-up, communication and reporting for you; you approve each step.

Watches the program Applications, matches, meetings
Does the work Matches, e-mails, rules, reports, certificates
You approve One click; nothing changes without you

Where you find Mentornity AI

Mentornity AI inside your admin panel

Ask it in the side panel, or press a "with AI" button where you work.

Ask in your own words

It reads the live program and answers with numbers and charts.

"With AI" buttons where you work

Applications, quiet matches, health, reports and forms each have one.

Nothing runs until you send

A button opens a ready request; credits are used only when sent.

Write with AI in the editors

E-mails, posts, events and agreements get a draft you can insert.

Ask in your own words

It reads the live program and answers with numbers and charts.

"With AI" buttons where you work

Applications, quiet matches, health, reports and forms each have one.

Nothing runs until you send

A button opens a ready request; credits are used only when sent.

Write with AI in the editors

E-mails, posts, events and agreements get a draft you can insert.

Matching proposals

Mentor proposals for every startup

It scores mentors with your matching rules and names the rules each one meets.

Three mentors, with the reasons

Ask for mentors per startup; each pick comes as a card.

Rules that separate, proposed

It spots a rule every pair passes and proposes a sharper one.

Expertise gaps in the pool

It compares what startups need with what mentors offer.

Three mentors, with the reasons

Ask for mentors per startup; each pick comes as a card.

Rules that separate, proposed

It spots a rule every pair passes and proposes a sharper one.

Expertise gaps in the pool

It compares what startups need with what mentors offer.

Quiet matches

Find quiet matches before they lapse

It lists active matches by how long they have been silent.

Most silent first

Never met, lapsed or nothing booked, each with its days of silence.

A restart e-mail you approve

It drafts a short e-mail; nothing is sent until you approve.

Most silent first

Never met, lapsed or nothing booked, each with its days of silence.

A restart e-mail you approve

It drafts a short e-mail; nothing is sent until you approve.

Application screening

Screen applications by your own rule

Write the rule once in your words; it checks every new application against it.

Mentor availability

Mentor time for the next two weeks

See who has open slots, who carries the load and whom to ask.

Meeting tracking

Meetings per startup, held and booked

Ask which startups have nothing booked and get the list, not only a count.

Term report with charts

An end-of-term report with charts

It writes the report from live data and explains any report you open.

Numbers, charts and two proposals

Reach, meetings per month and founder ratings for your sponsors.

Any report, explained

Open a report and press "Explain with AI" for its findings.

Numbers, charts and two proposals

Reach, meetings per month and founder ratings for your sponsors.

Any report, explained

Open a report and press "Explain with AI" for its findings.

Scheduled tasks

Checks that run on a schedule

Up to five tasks per program; each run brings its proposals as cards.

A new program

Build the next program with AI

Answer a few questions in New program; its assistant then finishes the setup.

Requests to try

Requests to start with

Type one in the side panel; it reads the data itself.

"Which startups have no meeting booked?" A list with meetings held and booked
"Suggest three mentors for each unmatched startup." Match cards, each with its reasons
"Check the new applications against my rule." Accept and archive cards
"Is there enough mentor time for the next two weeks?" Open slots and mentor load
"Prepare the term report for our sponsors." Numbers, charts and two proposals

Advice

How to get good answers

A few habits make its proposals fit your program.

Fill in the program profile Goals, audience and tone shape every answer
Give your screening rule once It is saved and used on every check
Ask for a chart when you compare Bar, line and pie charts from live numbers
Start with one weekly check Add tasks one at a time, five at most
Approve or reject every card Nothing changes until you decide

Privacy and cost

Privacy, credits or your own key

The model sees aliases, and you pay with credits or your own Anthropic key.

Names stay with you

The model sees participants under aliases, never their contact details.

Credits or your own API key

With your Anthropic key, Anthropic bills you and no credit is used.

Names stay with you

The model sees participants under aliases, never their contact details.

Credits or your own API key

With your Anthropic key, Anthropic bills you and no credit is used.

Claude connector and MCP

Mentornity inside Claude for your participants

Mentors and founders ask Claude in plain words; it books, plans and replies in Mentornity.

Claude, or any MCP assistant

They add the Mentornity connector and sign in with their usual account.

Meetings, hours and messages

Claude finds free slots, books, reschedules and drafts replies on request.

A preview before every change

Claude shows what will change; nothing happens until they say yes.

The program team decides

Claude sees only what the member sees; you can turn it off.