Built for operators · works insideChatGPT&Claude

Hand it off once. It stays handled.

Dorinda lives inside the ChatGPT or Claude you already use. It tracks every thread you're waiting on, chases the people who owe you answers — staged for your sign-off — and briefs you in seconds, not dashboards.

14-day free trial, no card · $15/mo · your data never trains a model.
No new app — you just ask
what's at risk this week?
Three threads: Legal's redlines (due Thu, nudged twice, no reply — escalate?), the Q3 vendor quote (5 days quiet), and Alex owes you the board one-pager (promised Monday).
escalate legal. push vendor to next week. leave alex.
nowLegal — escalation drafted, your OK
Mon 9:00Vendor quote — follow-up moved
due MonBoard one-pager — watching Alex
The problem

You became the tracking system. That was never the job.

The asks go out fine. It's everything after that lands back on you — the chasing, the checking, the remembering who owes what.

Your inbox is where follow-ups go to die.

You sent the ask. They didn't answer. Unless you personally remember to chase, the thread is dead — and you're carrying a dozen of them at any moment.

Another dashboard you have to open.

Every tracking tool works only as long as you keep visiting it. Status you have to go get isn't status — it's homework.

Re-briefing your tools.

Every AI chat starts from zero. You re-explain the org, the deal, the context — again — to get one useful answer.

Delegation without follow-through is hope.

“Handle it” only works if somebody confirms it got handled. Today that somebody is you, at 11pm, going through your sent folder.

The EA gap.

An assistant who chases, remembers, and reports is a luxury most operators run without. The follow-through work still exists — it's just landing on you.

How it's different

Status in seconds. Chasing at volume. Your rules.

Dorinda works inside the AI you're already talking to — and it runs the follow-through layer: who owes what, what's slipping, what needs your sign-off.

Ask “what's at risk?” — get an answer, not a dashboard

One typed question, one typed answer: what's overdue, what's quiet too long, what needs you. No tabs, no filters, no ritual. The briefing comes to the chat you're already in.

Asked and answered — in the chat
status on the offsite?
Venue confirmed. Catering owes a quote since Tuesday — I nudge tomorrow unless you want it today. The agenda draft is waiting on you.

It chases people, so you don't have to

Every thread you're owed — a quote, a signature, a deliverable — gets tracked and nudged, politely and persistently, with escalation drafted for your sign-off. A dozen threads cost you zero attention until one actually needs you.

Contract signature
waiting on outside counsel — 8 days
Tue 9:00Nudged — politely
Thu 9:00Re-nudged, cc added
Fri 8:12Escalation drafted for your OK

Set the policy once. It holds.

“Never send external email without my sign-off.” “Anything from the board reaches me same day.” Dorinda remembers your rules and applies them — approvals staged, never assumed.

new rule: nothing external goes out without my OK
Policy set — sends now stage first
Follow-up → vendor
“Hi — circling back on the Q3 quote. Could we have it by Friday?”
EditApprove & send

Same AI. Now it doesn't let go.

 
Just your AI
+ Dorinda
Chases the people you're waiting on
Keeps going until it's done — not one ping
Briefs you without being asked twice
Enforces your approval rules
One memory across ChatGPT & Claude
Remembers what you told it
Can remind you once, later
Answer-first

“What's at risk?” is a complete sentence.

You shouldn't have to assemble your own status meeting. Ask in four words, get the answer in five seconds, decide, move on. That's the whole interaction.

Asked, tracked, chased, closed — without a single dashboard.
Who's using it

How operators run on it.

The Monday sweep

Ten minutes before the leadership sync: “what moved, what's stuck, what's waiting on me?” Three typed answers, zero tabs opened. The meeting starts already briefed.

The chase ledger

A COO with eleven open asks across four teams. No spreadsheet, no standing reminder to check one. Dorinda nudges each owner on schedule and escalates the two that go quiet — drafts staged for sign-off.

The standing rule

“Nothing external without my OK” — set once, in one sentence, in January. Every follow-up since has staged, waited, and gone out exactly when approved. The rule never needed repeating.

  • Your data never trains a model
  • No server-side AI — a deterministic record
  • You own it and can export anytime
Pricing

Less than one interrupted hour costs you.

Full access, no card needed to start. Go Personal for just you, or bring the household onto Family — it lives in the AI you already pay for.

PersonalMost loved$15/mo

14-day free trial, then $15/mo — or $144/yr (~$12/mo, save ~20%).

Unlimited tracked threads & follow-ups
Connect your own Claude or ChatGPT via MCP
Escalating reminders, quiet-hours aware
Email drafts staged for your sign-off
Start free trial
FamilyUp to 5$40/mo

14-day free trial, then $40/mo — or $444/yr (~$37/mo), up to 5.

Everything in Personal — your household runs on it too
The shared Household — calendar, lists & projects
Roles — Owner · Adult · Teen · Assistant
Shared vs. private items, so everyone keeps their own
Start free trial
See full pricing & plan comparison →
Questions

The honest answers.

Is this a replacement for my EA?

No — it's leverage. Dorinda does the mechanical follow-through: tracking, chasing, nudging, briefing. Judgment, relationships, and the work itself stay with you and your people. If you have an assistant, Dorinda makes them faster; if you don't, it closes part of that gap.

Does it send email on its own?

Never by default. Nudges and follow-ups are drafted and staged for your sign-off. You can set your own rules — and rules only ever get tighter without you, not looser.

How fast is a status answer?

Seconds, in the chat you already have open. “What's at risk?”, “what's waiting on me?”, “status on the deal?” are all complete questions — the answer is typed back, not linked to a dashboard.

Do I need to download a new app?

No. Dorinda works through the ChatGPT or Claude you already use — you just talk to it like normal. Nothing new to install, learn, or check.

How much does it cost?

$15/mo for Personal, $40/mo for Family (up to 5). Every plan starts with a 14-day free trial — full access, no card needed.

Do I need a separate AI subscription?

Dorinda connects to the ChatGPT or Claude you already pay for — bring your own. It gives that AI the memory and follow-through to run your threads, with no lock-in.

Stop being your own tracking system.

“Dorinda's got it.”
Start free trial — no card, under a minute