Still asking “Any update?”

Know where every piece of work stands, and what should happen next.

Inistate gives your people, AI, and automations one shared operational workflow, so everyone knows what’s happening, what can happen next, and who needs to act.

Free forever for small teams. No credit card required.

Running operations at Omni-Plus, Lavender, PC Dreams and RentWorks.

One record. One state. Only the activities that state permits. Every actor on the same history.

Everyone’s doing their part. Nobody has the full picture.

The lead gets enriched. The ticket gets routed. The approval happens in a chat thread. Every automation runs fine, and a manager still becomes the human router between them.

Today you ask

“What’s happening with this?”
“Who has it?”
“Are we waiting on someone?”
“Did the AI already do it?”

With one workflow

Here is the state it’s in.
Here is who owns it.
Here is what it’s waiting for.
Here is what acted, and when.

One primitive, not a pile of conventions

Inistate is an operational workflow platform. Every job, case, request or claim on it runs on the same four-part contract, whoever moves it — which is what makes the status trustworthy.

State Activity (Form) State

and every one of them is written to history

Without it the status lives in a spreadsheet column, a chat thread and somebody’s memory — and they disagree.

Without it anyone can set any status at any time, so a record can say Completed while the work is half done.

Without it work moves on with the photo, the reason or the amount missing, and somebody chases it afterwards.

Without it you have logs that say a workflow ran, not who approved what, or why.

See the handoffs

People, AI and automation. One job, together.

Open an example and follow the work from request to completion. Different strengths, one shared history.

Field OperationsFrom a reported fault to a finished job.
Field technician reviewing a job on a tablet
One record, start to finishJOB 1042 · Air conditioner leaking

Example workflow Choose a step to see the handoff

AIStep 1 of 5

Triage the request

Requested → Ready to assign

AI reads the fault report, identifies the equipment and suggests a priority.

A clear job brief, ready for dispatch.
AutomationStep 2 of 5

Dispatch the job

Ready to assign → Assigned

Assignment rules route the job to the available field team and send the customer an appointment notification.

The technician and customer know what happens next.
PeopleStep 3 of 5

Complete the repair

Assigned → Work completed

The technician fixes the leak and submits site photos, time spent and a completion note.

Proof of work stays with the job.
AIStep 4 of 5

Check the evidence

Work completed → Evidence missing

AI checks the submission and finds the completion photo missing, so it sends the job back with exactly what is needed. Sign-off is not offered until it is there.

The exception has an owner and a reason, not a message in a group chat.
PeopleStep 5 of 5

Resubmit and sign off

Evidence missing → Closed

The technician adds the photo from the site. The supervisor reviews the evidence and signs off, and automation sends the completion notice to the customer.

One closed job, with the detour on the record.
Software Development IssuesFrom a reported problem to a verified release.
Software development team investigating an issue together
One record, start to finishISSUE 1847 · Checkout fails after address update

Example workflow Choose a step to see the handoff

PeopleStep 1 of 5

Report the issue

New → Reported

A support specialist records what happened, who is affected and the steps needed to reproduce the problem.

The development team receives a clear issue with evidence attached.
AIStep 2 of 5

Triage the issue

Reported → Ready for development

AI summarizes the report, identifies the affected area, checks for related issues and suggests a severity for the product owner to review.

The issue arrives organized, with likely impact and context.
PeopleStep 3 of 5

Investigate and fix

Ready for development → In review

A developer reproduces the failure, implements a fix and links the code change and test evidence to the issue.

The proposed fix and its reasoning stay with the issue.
PeopleStep 4 of 5

Review and verify

In review → Ready to release

A teammate reviews the change while QA verifies the original scenario and records the test outcome.

Human review confirms the fix is safe to release.
AutomationStep 5 of 5

Release and notify

Ready to release → Resolved

Entering Ready to release triggers the deployment pipeline. When it reports success, automation updates the issue and notifies the reporter.

The issue closes with the deployment and notification results in one history.
Service & SupportGive every request a clear path to resolution.
Support specialist helping a customer at her desk
One record, start to finishTICKET 2086 · Unable to access account

Example workflow Choose a step to see the handoff

AIStep 1 of 5

Understand the issue

New → Triaged

AI summarizes the customer’s message, categorizes the issue and suggests troubleshooting steps.

The request arrives with useful context.
AutomationStep 2 of 5

Route to the right team

Triaged → Assigned

Routing rules assign the ticket to account support and start the response-time reminder.

A named team owns the next action.
PeopleStep 3 of 5

Resolve the exception

Assigned → Resolved

A support specialist verifies the account, restores access and records the solution.

The customer has a solution, and the team has a record.
AutomationStep 4 of 5

Ask for confirmation

Resolved → Awaiting confirmation

Automation sends the resolution to the customer and schedules a follow-up reminder.

The ticket stays visible while the team waits.
PeopleStep 5 of 5

Confirm it works

Awaiting confirmation → Closed

The customer confirms access is restored. The specialist closes the ticket with the resolution attached.

The conversation and its outcome stay together.
Work Order ManagementMove a request through approval and delivery.
Two colleagues reviewing operational information with a tablet
One record, start to finishWO 0318 · Replace a damaged pump

Example workflow Choose a step to see the handoff

AIStep 1 of 5

Prepare the request

Requested → Pending approval

AI turns the maintenance note into a work order with the scope, supporting details and a cost estimate for review.

The approver has a structured request to assess.
PeopleStep 2 of 5

Approve the work

Pending approval → Approved

The manager reviews the scope and estimate, confirms the budget and records the decision.

A person owns the spending decision.
AutomationStep 3 of 5

Schedule and notify

Approved → Scheduled

Automation assigns the approved order to the maintenance team and sends the schedule to everyone involved.

Work is scheduled with a clear owner.
PeopleStep 4 of 5

Carry out the work

Scheduled → Completed

The team replaces the pump and records the parts used, completion notes and inspection photos.

The work order holds the delivery evidence.
PeopleStep 5 of 5

Verify and close

Completed → Closed

The manager checks the completed work and signs off. Automation updates the asset record and notifies the requester.

The request, approval and completed work form one history.
Auditors & AccountantsFrom missing client documents to a reviewed file.
Two accountants reviewing client documents together at a desk
One record, start to finishREQUEST 0426 · Year-end supporting documents

Example workflow Choose a step to see the handoff

PeopleStep 1 of 5

Define what is needed

New → Awaiting documents

The engagement lead lists the statements, invoices and supporting schedules needed from the client, with an owner and due date.

The client and team share one clear document request.
AutomationStep 2 of 5

Collect and remind

Awaiting documents → Documents received

Automation sends the request, reminds the client about outstanding items and updates the request when the requested uploads arrive.

Documents stay linked to the request, ready for checking.
AIStep 3 of 5

Prepare the review

Documents received → Ready for review

AI organizes the uploads, matches them to the checklist and flags possible gaps or inconsistencies for the reviewer.

The reviewer receives an organized file with points to check.
PeopleStep 4 of 5

Review and sign off

Ready for review → Reviewed

The auditor or accountant checks the source documents, resolves questions with the client and records their review decision.

Professional judgment and the review notes stay with the file.
AutomationStep 5 of 5

Notify the engagement team

Reviewed → Closed

Entering Reviewed triggers a notification to the engagement owner. Once the notification result is reported back, the request closes with the reviewed file linked.

The engagement team knows the file is ready, with every handoff recorded.
Property AgenciesFrom a property enquiry to a completed viewing.
Bright property entrance lobby with floor-to-ceiling windows
One record, start to finishENQUIRY 0732 · Apartment viewing request

Example workflow Choose a step to see the handoff

AIStep 1 of 5

Understand the enquiry

New → Ready to assign

AI reads the enquiry and extracts the preferred location, budget and viewing availability, flagging missing details for follow-up.

The agent receives a clear brief with the original enquiry attached.
AutomationStep 2 of 5

Assign an agent

Ready to assign → Assigned

Routing rules assign the enquiry to the agent responsible for the listing and notify them of the new request.

A named agent owns the next action.
PeopleStep 3 of 5

Arrange the viewing

Assigned → Viewing scheduled

The agent confirms that the property is available, agrees a time with the prospect and records the appointment details.

The prospect, property and appointment stay on one record.
AutomationStep 4 of 5

Send the reminder

Viewing scheduled → Reminder sent

The scheduled state triggers an appointment reminder with the address and contact details. The reminder result is reported back to the enquiry.

The agent can see that the viewing reminder was sent.
PeopleStep 5 of 5

Record the outcome

Reminder sent → Viewing completed

After the viewing, the agent records feedback and agrees the next action: send more details, arrange another viewing or prepare an offer for review.

The viewing is complete, and the next follow-up has a clear owner.

Different strengths. One workflow.

People make decisions. AI interprets information. Automation connects systems. All three move the same work forward, under the same rules.

Choose a card to see how each connects to Inistate.

People work in the app — web and mobile

The Inistate web app, listing records with their current state The Inistate mobile app home screen

They see the record, the state it is in, and only the activities that state permits — so the form is filled in before Complete Job is available.

AI works in its own client — over MCP

Claude ChatGPT or any MCP client

The aircon repair at Unit 12B is done — here is the photo.

Inistate · tool call

submit_activity
record
JOB 00042
activity
Complete Job
form
completion photo

Done — JOB 00042 is Completed. I could not have skipped the completion photo: the activity requires it.

How many repairs are still open past their SLA this week?

read · list_entries Six, and four are with the same contractor. I can put that in a report with the time in each state.

The agent gets no private route: it reads what the state permits and submits the same activity, through the same form. The same connection answers questions and writes the report — from the live records, not a stale export.

Automation works on its own canvas — one node is Inistate

Through the n8n, Make and Zapier apps, or straight at the REST API and webhooks.

n8n Make Zapier or the REST API and webhooks

1 Into the workflow — your automation performs the activity

Trigger Photo received
Inistate node Complete Job completion photo JOB 00042 · In Progress → Completed

2 Out of the workflow — the activity and the state it lands in are triggers of their own

Inistate trigger Complete Job performed state changed · In Progress → Completed
Slack Post to #ops
Google Calendar Book the follow-up

One node performs the activity the state permits — send it without the completion photo and it fails, instead of writing a status nobody can trust. And because the activity and the new state are recorded, they are things your next automation can wait for.

Three channels. One record: JOB 00042 · Aircon repair · Unit 12B, In Progress → Completed, on one history.

People judge. AI reasons. Automation executes.

Works with what you already run

Automate anywhere. Govern the workflow in Inistate.

An action moves the work forward. The new state starts the next automation. Follow each handoff on the same record.

People

Web Mobile

Your team, on web or mobile

AI

or any MCP client

Automation

Keep executing where you already do

InistateOne workflow. Every actor. Same rules.

Example · JOB 00042

Aircon repair · Unit 12B

In Progress
StateIn Progress
Activity · FormComplete Job
Next stateCompleted

The repair is finished. Include a completion photo to close out the work.

3. The new state triggers an automation

Waiting for state change
When the record entersCompleted
n8nNotify the tenant

Send the completion summary and photo to the tenant.

Complete the activity above to see the automation run. A blocked activity does not trigger it.

Illustrative example. A trigger starts the automation; its outcome is recorded when the connected tool reports back.

4. One shared history

See how the work got here.

Who acted, what they submitted and how the state changed. Follow the handoffs on the same record, including automation results reported back.

Person Mei logged the repair request from a tenant photo.
AI An agent suggested Rafiq based on availability.
Automation n8n sent the tenant a reminder for the visit.

And the tools you already connect through them

Airtable Asana PostgreSQL MySQL MongoDB MariaDB

Thousands more reach the same workflow through Zapier, Make and n8n. See how the connection works →

Operations already running on Inistate

Different operations, different processes, one place the work actually lives.

Omni-Plus System Lavender Confectionery & Bakery PC Dreams Group RentWorks Strategy
“Inistate’s flexibility and robustness are impressive where it covers our customized internal business process for Lavender’s fast-growing expansion.”
Tam Yong Wen · Chief Executive Officer, Lavender Confectionery & Bakery

ISO/IEC 27001:2022 certified Read the case study → Trust and compliance →

Is this an Inistate process?

Inistate is not for every automation. It is for work that outlives a single run. Answer these about one process you have in mind.

Is there a thing that needs tracking: a job, case, request, order, claim, or approval?
Does it move through several meaningful states?
Do people sometimes review, assign, approve, correct, or complete it?
Do AI or automations also act on it?
Does somebody regularly ask what the status is?
Does the team need to know what can happen next?
Does it matter who did what?
Does it last longer than one automation run?

Mostly yes

That is an Inistate process. Give it one workflow and the questions above stop being questions.

Mostly no

A form to an email to a spreadsheet row, a scheduled sync, or one isolated approval does not need this. Your automation tool already has it covered.

Questions, answered.

What it is, what it is not, and how it fits with the tools you already run.

One workflow per process, not one for the whole company. A job, a claim, a ticket or an onboarding each gets its own: the states it moves through, the activities each state permits, the form each activity requires, and one history of who did what. Start with the process that generates the most “any update?” messages. Add the next one when it has earned its place.

No, and it should not. Those tools execute: they move data, call systems and run the steps that never need judgment. Inistate holds the workflow around them — the state the work is in, what may happen next, who may act, and what already happened. Your automation performs an activity through the Inistate app on n8n, Make or Zapier, or straight at the REST API and webhooks, and the state it lands in can trigger the next automation. Many automations. One governed workflow.

Through MCP. Inistate is in the Claude and ChatGPT connector directories, and any other MCP client connects the same way. Once connected, the agent has no private route into your data: it can only perform the activities the current state permits, with the form each one requires, within the permissions you give it — never wider than the person it acts for. An activity your rules reserve for a person, like approving spend, stays with a person. It reads the same way it writes — ask it how many jobs are past their SLA this week and it answers from the live records, not an export.

It stops and asks. Past a confidence threshold the agent pauses for a person instead of guessing, and the question and the answer stay on the record. Every action it does take carries its reasoning, its sources, its confidence and the model used, on the action itself rather than in a separate log. So when something needs a second look you can see exactly what happened and why — and because the workflow rules apply to the agent too, it could not have skipped a required step to get there.

No. Describe the process in a sentence, forward the message you usually get, or photograph the form you use today, and Inistate turns it into a structured starting point — states, fields, owners and history — rather than a blank canvas. You shape it from there, without code. Bring one process, not the org chart; the rest of the operation joins one workflow at a time.

Free forever, no card. It includes 2 users, plus an actor seat that connects the integrations and AI clients and can go to a third person, an automation or an AI agent; 500 active records plus 500 read-only; 1 GB of files; 30 days of change history; and unlimited modules, guests, documents, reports and built-in automations for notifications, reminders and alerts. It also comes with 1,000 AI credits, one-time, which is about enough for the AI to build one complete workflow with you. When you need more people or connections, add up to 5 more actor seats at US$8 a month each. If you outgrow a limit we warn you and suggest the next plan — we never refuse to take in your work. Details on pricing; for teams of 50 or more, see enterprise.

In credits, shared by the whole workspace, and they never expire. Plans buy the capacity — users, records, storage, history — and AI is metered separately, so a quiet month costs nothing extra and you are never paying a monthly fee for AI you did not use. Ordinary work such as logging a job from a photo, chasing an approval or drafting a document costs far less per turn than asking the AI to design or analyse something. Top up with a pack whenever you want more.

Yes to both. Inistate runs on the web and on iOS and Android, so a technician completes the job and attaches the photo from the site, and the record updates for everyone at once. For people outside your team — a customer confirming the fix, a client uploading year-end documents — you share a form or a record with them as a guest. Guests are unlimited, even on the free plan.

Those tools give you a place to put work, and increasingly an agent to help with it. But your team is still the workflow engine: the rules live in columns, buttons and habits. In Inistate the workflow is the product: the state, the permitted activities, the required forms and the history are what people, AI and automations all act through. Jira is built around tickets, sprints and epics for software teams; Inistate is built around states, activities and forms for any operation. App builders like Glide or Jotform build you an app and hand it back; Inistate stays inside the workflow and runs it with you.

Do — through Inistate. On its own, a chat gives you an answer in a window and forgets it when the conversation ends. Connected to a workflow, the same assistant performs activities on real records, within the rules, with every action on the history, and it is still tracking the work when the day ends. Keep the assistant you already use. Give it something that persists.

Inistate is ISO/IEC 27001:2022 certified, and the architecture is protected by US and Singapore patents. Because who did what, with which input, from which state, is written on the record itself, the audit trail is a by-product of the work rather than something reconstructed from exports and screenshots under deadline. For regulated teams, those fields are evidence your compliance team can map to obligations such as EU AI Act Article 14, FINRA Rule 3110, the Colorado AI Act and HIPAA audit controls. See trust and compliance and the privacy policy.

Work that fits inside a single automation run. A form that emails a spreadsheet row, a nightly sync, one isolated approval — your automation tool already handles those, and a workflow around them is overhead. Inistate is for work that outlives the run: a job, case, request, order or claim that moves through several states, that people and systems both touch, and that somebody keeps asking about.

One workflow. Every actor. Same rules.

Stop chasing updates.

Bring one process. Put your team, your AI, and your automations on it. See where it stands without asking anyone.

Free plan, no card ISO/IEC 27001 certified Works with n8n, Make and Zapier iOS & Android