51 client flows that work on hard days.
Intake, follow-up and check-ins that still work when someone is overwhelmed and you are out of hours. Browse the options, then we build the ones that fit your practice.
Tell me the problem. I’ll point you to the fix.
Five things that come up again and again for practices, and exactly which flows solve each one. Or skip ahead and browse all 51 below.
“New clients fall through the cracks between booking and their first session.”
Client Intake System
Intake form, confirmation, practitioner notification, a structured sheet, the pre-session brief and Drive organization.
“You’re still the one chasing every lead, reminder and review request yourself.”
Client Journey Automation
Lead capture, onboarding, scheduling, reminders, intake, follow-up, rebooking and review requests.
“Clients need support between sessions, and you can’t be reachable around the clock.”
Nervous-System-Friendly Client Portal
A state check-in, regulation menu, personalized resources, session preparation, recap and tracking over time.
“You start every day with no idea what actually needs your attention.”
Practitioner Command Center
One screen for appointments, missing intake, leads, follow-ups, flags and tasks.
“Routine writing and sorting still eats hours you’d rather spend with clients.”
AI Admin Layer
Email drafts, session summaries, recap drafts, content repurposing, FAQ, sorting and extraction.
Four rules, no exceptions.
One next step
The person does the least possible and gets the most useful next step. Never a list.
A way out
Every flow shows a route to a human or emergency resource, and a “not helping” option.
A person decides
Anything sensitive, from safety to interpretation, is reviewed by a person you name.
AI drafts, never decides
Rules and thresholds stay in plain logic. AI only summarizes, drafts and sorts, and never judges safety.
Pick the flows that fit your practice.
Universal check-in
Give the person one recommended next step matched to their current capacity.
How it works
- Starts when
- Client opens the check-in link, app or pre-session form.
- What runs automatically
- Logs the response, sends the matching resource and notifies you when something is flagged.
- Where AI is used
- None required. Optional: rewrite resource text in the client’s tone.
- Where a person is in charge
- A person reviews every safety-flagged response within an agreed window.
30-second crash
Provide instant help with zero thinking required.
How it works
- Starts when
- A persistent “I need help now” button on every page, message and portal screen.
- What runs automatically
- Four buttons on one static page, with an optional tap to message you (clearly not monitored live).
- Where AI is used
- None. Predictability is the feature.
- Where a person is in charge
- You approve the exact wording once.
Wired flow
Lower activation enough to decide whether to solve something or slow down first.
How it works
- Starts when
- Client’s state is wired, or they tap “come down.”
- What runs automatically
- A 90-second guided path, with the supports they chose and rated logged.
- Where AI is used
- None in the moment. Later, can summarize which supports the client rates as helpful.
- Where a person is in charge
- You review the resource content and any escalation trigger.
Shutdown flow
Restart small movement without demanding energy the person doesn’t have.
How it works
- Starts when
- Client selects numb, frozen, blank, can’t start or exhausted.
- What runs automatically
- One tiny action at a time. Three shutdown check-ins in a row raises a soft flag in your brief.
- Where AI is used
- None.
- Where a person is in charge
- You decide what a repeated pattern means. The system does not interpret it.
Sensory overload
Reduce the strongest controllable input and lower communication demands.
How it works
- Starts when
- Client selects “too much going on” or reports noise, light, touch, smell, crowding, temperature or clutter.
- What runs automatically
- Saves each client’s sensory preferences and uses them next time.
- Where AI is used
- None.
- Where a person is in charge
- You confirm the sensory preferences are captured correctly at the next visit.
High activation and rage
Increase the space between activation and action. The goal is not to make anger disappear.
How it works
- Starts when
- Client selects angry, furious, about to send something, or high activation.
- What runs automatically
- A pause, a distance action and a timing decision, with an optional 30-minute reminder to revisit a draft. Never sends anything for the user.
- Where AI is used
- None while activated. Drafting help only after activation drops.
- Where a person is in charge
- Any mention of threats, harm to others or abuse stops the automated flow and goes to a person.
Conflict flow
Turn a charged situation into a clear, calm message, or a decision not to send one.
How it works
- Starts when
- Client reports a conflict with a partner, colleague, parent or client.
- What runs automatically
- Breaks the situation into what happened, how it was read, feeling, need and request, then saves a draft only.
- Where AI is used
- Structures the notes and writes a short draft. No blame, no diagnosis, no advice. Safety language returns a flag instead.
- Where a person is in charge
- The user reviews and sends. You only see summaries, with consent.
Repair after rupture
Help someone repair a strained relationship with actions, not just words.
How it works
- Starts when
- A client, parent, partner or team member asks how to make it right.
- What runs automatically
- Saves a repair commitment and schedules one check-in after seven days.
- Where AI is used
- Optional wording help, kept to the user’s own facts.
- Where a person is in charge
- In practitioner-client rupture, the practitioner owns the repair conversation.
Co-regulation for supporters
Help practitioners, parents, partners and support workers respond to someone who is dysregulated.
How it works
- Starts when
- A supporter opens the card or asks how to help someone in distress.
- What runs automatically
- A one-page card with a first question, four options and an exit question. Optional log of what helped.
- Where AI is used
- None.
- Where a person is in charge
- Any safety concern about the person or supporter moves to a person-led safety route.
Parent and child
Give a parent one useful step in a hard moment, and build a child-specific profile over time.
How it works
- Starts when
- Parent taps “my child is struggling.”
- What runs automatically
- An in-the-moment card, a single later log prompt, and a profile of what has helped.
- Where AI is used
- Summarizes logs into patterns, phrased as observations.
- Where a person is in charge
- A qualified professional reviews any developmental or safety concern.
Decision fatigue
Get someone unstuck from too many choices without forcing perfection.
How it works
- Starts when
- Client reports too many decisions, being stuck choosing, or option overload.
- What runs automatically
- Four taps lead to a default choice, or a deferral with a date and a reminder.
- Where AI is used
- Optional: suggests two or three defaults from the user’s own criteria.
- Where a person is in charge
- High-stakes decisions such as legal, medical or financial go to a qualified professional.
Task initiation and follow-through
Find which stage of a task breaks down, and give a fix that matches it.
How it works
- Starts when
- “I know what I need to do but can’t start.”
- What runs automatically
- Sorts the problem into see, start, sustain, finish or transition, then schedules a two-minute start reminder and a done check.
- Where AI is used
- Breaks a task into first physical actions, always from the user’s own words.
- Where a person is in charge
- A coach or practitioner reviews patterns monthly.
Burnout and overload diagnostic
Separate a capacity problem from a workload, environment or system-design problem.
How it works
- Starts when
- A client or employee reports burnout, exhaustion or overload at work.
- What runs automatically
- Scores eight drivers and produces a plain-language report naming the top two, with suggested conversations.
- Where AI is used
- Summarizes free-text comments, kept separate from the ratings.
- Where a person is in charge
- A manager, HR lead or practitioner reviews the report. Signals of depression, self-harm or serious health concern go to a person.
Morning capacity
Replace “what are your goals today” with “what is realistic today.”
How it works
- Starts when
- Daily message or app open.
- What runs automatically
- Builds a plan of no more than four items and adds the must-do to the calendar.
- Where AI is used
- Drafts the plan from the four answers, keeping the client’s words.
- Where a person is in charge
- A coach reviews patterns weekly.
Low-capacity day
Protect a person from doing more than they can, and from feeling bad about it.
How it works
- Starts when
- Capacity is low, or a check-in suggests exhaustion or overwhelm.
- What runs automatically
- Switches to a minimum day of about five items and hides everything else.
- Where AI is used
- None.
- Where a person is in charge
- Repeated low-capacity days over several weeks appear in your brief for human review.
High-capacity day
Use a good day well without spending tomorrow’s energy.
How it works
- Starts when
- Capacity is high.
- What runs automatically
- A capped priority list with a recovery block added to the calendar.
- Where AI is used
- Suggests ordering, based on the client’s stated priorities.
- Where a person is in charge
- None by default.
Transition support
Make switching between tasks, settings or roles less costly.
How it works
- Starts when
- A scheduled transition, such as leaving work, ending a session or starting a new task.
- What runs automatically
- Three timed messages (warning, buffer, next-step cue) and a two-line shutdown note.
- Where AI is used
- None.
- Where a person is in charge
- None.
Cycle tracking
Let a person see their own cycle and symptoms over time.
How it works
- Starts when
- A period start is logged, or a daily one-tap check-in.
- What runs automatically
- Rolling cycle length, a predicted window shown as a range, a calendar view and a monthly summary. Missing days stay missing, and the user can override any phase estimate.
- Where AI is used
- Summarizes the month in plain language, without prescribing.
- Where a person is in charge
- A practitioner reviews the summary before it becomes part of any care plan.
Symptom pattern detector
Show whether a symptom clusters at a repeatable time.
How it works
- Starts when
- At least three cycles of the person’s own data are available.
- What runs automatically
- A one-page pattern report after each cycle. With fewer than three cycles it shows progress, not a pattern.
- Where AI is used
- Writes a cautious summary with no diagnosis or causes, ending with “this is a pattern you may want to discuss with your clinician.”
- Where a person is in charge
- The client decides whether to share the report.
Symptom tracking and clinician export
Give the client and clinician clear, dated evidence instead of “I feel awful before my period.”
How it works
- Starts when
- A client or practitioner starts a symptom-tracking program.
- What runs automatically
- Daily ratings with safety questions built in, and a clinician-ready PDF table of date, symptom, severity and impact.
- Where AI is used
- Optional summary with no interpretation.
- Where a person is in charge
- A clinician, not the system, interprets the data. A practitioner reviews all safety flags.
Perimenopause mode
Avoid forcing a regular-cycle model onto an irregular pattern.
How it works
- Starts when
- Cycles are unpredictable, or predictions keep missing.
- What runs automatically
- Switches to a symptom-first rolling four-week view with manual override, plus a clinician export.
- Where AI is used
- Summarizes patterns without labeling a stage.
- Where a person is in charge
- A clinician interprets any staging or treatment question.
Postpartum check-in
Offer very light support and clear escalation for new parents.
How it works
- Starts when
- Scheduled check-ins at intervals you set.
- What runs automatically
- A one-screen check-in, with a same-day practitioner alert on any flag.
- Where AI is used
- None for safety. Optional draft replies that a person reviews.
- Where a person is in charge
- A trained person reviews every flag the same day, with an agreed after-hours plan.
Appointment prep
Help the client walk into an appointment prepared and heard.
How it works
- Starts when
- An appointment is booked with any clinician.
- What runs automatically
- A one-page prep sheet built from their own entries, sent two days before.
- Where AI is used
- Drafts the summary from the client’s own entries, adding nothing they did not say.
- Where a person is in charge
- The client approves before sharing.
Pre-session brief
You walk into the session knowing the client’s state, main concern and any flags.
How it works
- Starts when
- A form is submitted after booking, or a set number of hours before the appointment.
- What runs automatically
- Normalizes the answers, applies your written rules, emails a one-page brief and archives it, with duplicate handling.
- Where AI is used
- Optional two-sentence summary of free text. Rules and flags stay deterministic.
- Where a person is in charge
- You read every brief and any priority flag before the session.
First-visit intake
Gather deeper context once, then never ask the same thing twice.
How it works
- Starts when
- First appointment booked.
- What runs automatically
- Save-and-resume intake, a client folder and a one-page Client Snapshot.
- Where AI is used
- Drafts the snapshot, labeled client-reported.
- Where a person is in charge
- You review the snapshot and consent before the first session.
Follow-up visit
Build continuity without repeating the whole intake.
How it works
- Starts when
- A return appointment is booked.
- What runs automatically
- Pre-fills what is known and highlights only what changed.
- Where AI is used
- Summarizes change since the last session.
- Where a person is in charge
- You review the change summary and flags.
Between-session check-in
Give clients a small, easy way to stay connected without creating an obligation.
How it works
- Starts when
- Scheduled message at the interval you set.
- What runs automatically
- Five buttons, routing and logging, added to your next brief.
- Where AI is used
- None for routing. Optional summary of notes.
- Where a person is in charge
- A person reads every “struggling” or “need support” response within the agreed window.
No-response flow
Make sure automation reduces burden instead of adding obligation.
How it works
- Starts when
- No reply after a set interval.
- What runs automatically
- One soft message, then a hard stop. Never escalates tone.
- Where AI is used
- None.
- Where a person is in charge
- You are notified only if you chose that in advance.
Safety flag review
Make sure a person sees every safety flag quickly, with a written plan for what happens next.
How it works
- Starts when
- A safety answer is yes, a keyword is detected, or a client uses the “I need support” button.
- What runs automatically
- Alerts the designated person on a channel they watch, logs it, and alerts a backup if nobody acknowledges.
- Where AI is used
- None. AI is never used to decide whether someone is safe.
- Where a person is in charge
- A trained person owns every decision, using a protocol you define with your own professional guidance.
AI session summary
Save time on notes while keeping client facts and AI interpretation clearly separate.
How it works
- Starts when
- You submit structured notes after a session.
- What runs automatically
- A six-section draft saved to the client folder.
- Where AI is used
- Summarizes and organizes, labeling each statement client-reported or AI-inferred. No diagnosis or advice.
- Where a person is in charge
- You read and edit before it becomes part of any record.
AI client recap
Give the client a plain-language recap and a next step they chose.
How it works
- Starts when
- You approve the session summary.
- What runs automatically
- Drafts a short recap and queues it for your review.
- Where AI is used
- Rewrites the approved summary in warm plain language, using only the client’s words.
- Where a person is in charge
- You approve every recap before it is sent. Not approved in 48 hours means nothing is sent.
No-show and cancellation recovery
Bring people back without shame and protect the schedule.
How it works
- Starts when
- A no-show or late cancellation is recorded.
- What runs automatically
- One kind message with a one-tap rebooking link, then stop.
- Where AI is used
- None.
- Where a person is in charge
- You decide whether to reach out personally, especially if the client has been struggling.
Practitioner command center
One screen that shows what needs your attention today.
How it works
- Starts when
- Dashboard opened, or a morning digest.
- What runs automatically
- Today’s appointments, missing intake, new leads, follow-ups and flags (flags first), plus a counts-only morning email.
- Where AI is used
- Optional plain-language summary of the day.
- Where a person is in charge
- You own every action shown.
Lead capture and qualification
Turn an inquiry into a booked conversation with the least effort on both sides.
How it works
- Starts when
- Website form, DM or email inquiry.
- What runs automatically
- Acknowledges within minutes, checks fit, logs the lead and sends a calendar link. Non-fits get a kind referral.
- Where AI is used
- Optional classification of free text. Drafts a reply for unusual messages.
- Where a person is in charge
- You review non-fit and unusual inquiries. Crisis language goes to a person at once.
Booking and confirmation
Make booking one step and confirmation calm and clear.
How it works
- Starts when
- An appointment is booked.
- What runs automatically
- Confirmation, intake link, reminders and calendar event.
- Where AI is used
- None.
- Where a person is in charge
- You approve the message wording once.
Onboarding
Give new clients an early win and a clear idea of what happens next.
How it works
- Starts when
- Payment or first booking.
- What runs automatically
- A welcome sequence of three to five short messages, one action at a time, with skip and pause.
- Where AI is used
- Optional personalization of the welcome, reviewed by you.
- Where a person is in charge
- You review the sequence copy once.
Reminders that respect capacity
Reduce no-shows without nagging.
How it works
- Starts when
- An appointment is scheduled.
- What runs automatically
- At most two reminders with reschedule and cancel buttons, respecting quiet hours and consent.
- Where AI is used
- None.
- Where a person is in charge
- None.
Rebooking
Make the next appointment the easy default.
How it works
- Starts when
- A session ends, or a set number of days pass without a future booking.
- What runs automatically
- A booking link in the recap and one follow-up, then stop.
- Where AI is used
- None.
- Where a person is in charge
- You choose which clients receive it.
Review and testimonial request
Collect feedback and reviews ethically.
How it works
- Starts when
- A set time after a completed program, only for clients with no open flags.
- What runs automatically
- Asks for a rating first, sends a review link only to positive ratings, and routes everyone else to private feedback.
- Where AI is used
- None.
- Where a person is in charge
- You decide whether to send to any client who has had a hard period.
Membership and community engagement
Keep members connected without requiring constant attention.
How it works
- Starts when
- Weekly schedule.
- What runs automatically
- One weekly message with one link, matched to each member’s chosen intensity.
- Where AI is used
- Optional summary of community threads for the host.
- Where a person is in charge
- The host reviews reports of harmful content and members in distress.
Program completion
Find and fix where people drop out.
How it works
- Starts when
- Program start.
- What runs automatically
- A weekly drop-off map and a soft nudge with a smaller option.
- Where AI is used
- Summarizes free-text exit feedback.
- Where a person is in charge
- The program owner decides changes to content.
Workplace or school wellbeing
Give an organization aggregate insight without watching individuals.
How it works
- Starts when
- Scheduled pulse survey.
- What runs automatically
- Anonymous collection and aggregate reporting only above a minimum group size.
- Where AI is used
- Summarizes themes with names and identifying details removed.
- Where a person is in charge
- A trained lead reviews reports. Any disclosure of harm is handled by a person.
Retreat logistics
Reduce guest anxiety and organizer admin before, during and after an event.
How it works
- Starts when
- Registration.
- What runs automatically
- Confirmations, packing list, day-of messages, a guest roster and a dietary summary.
- Where AI is used
- Optional summary of dietary and accessibility notes.
- Where a person is in charge
- The host reviews all medical and accessibility notes before the event.
Content repurposing
Turn your existing expertise into newsletters and posts without extra effort.
How it works
- Starts when
- A new article, talk or approved recap.
- What runs automatically
- A batch of drafts saved for review.
- Where AI is used
- Rewrites and reformats in your voice, adding no claims.
- Where a person is in charge
- You review for accuracy, privacy and health-claim risk before anything is posted.
Decision tree builder
Turn “I want to automate my intake” into a workable design before anything is built.
How it works
- Starts when
- Discovery call or audit interview.
- What runs automatically
- Design work, not software: a one-page map of trigger, input, decision, action, human checkpoint, output, exception and escalation.
- Where AI is used
- Organizes notes into the eight boxes and flags gaps. Never invents steps.
- Where a person is in charge
- You and the client both sign off on the map.
AI pattern detector
Surface possible patterns in structured data over time, without claiming cause.
How it works
- Starts when
- Several weeks or cycles of structured data exist.
- What runs automatically
- A weekly or monthly report with counts and cautious language.
- Where AI is used
- Wording only. The statistics are computed in code.
- Where a person is in charge
- You review the report before sharing.
Email and message drafting
Cut the time spent on routine replies while keeping your voice.
How it works
- Starts when
- A new message in a labeled folder.
- What runs automatically
- Saves a draft reply only. Messages about symptoms, medication or crisis are never drafted.
- Where AI is used
- Drafts from your approved answers.
- Where a person is in charge
- You edit and send every draft.
FAQ assistant
Answer common questions instantly without pretending to be a clinician.
How it works
- Starts when
- A visitor asks a question on your website or portal.
- What runs automatically
- Answers from approved content only, logs unanswered questions, and hands off to a contact route.
- Where AI is used
- Matches the question to your approved answers.
- Where a person is in charge
- You review the log weekly and update the content.
Classification and routing
Send each inbound item to the right person or workflow.
How it works
- Starts when
- A new email, form submission or message.
- What runs automatically
- Tags, assigns and notifies. Safety keywords are caught by rules first.
- Where AI is used
- Classifies the rest with a confidence value. It can escalate but never downgrade a safety flag.
- Where a person is in charge
- A person reviews low-confidence items and everything tagged clinical or safety.
Document extraction
Pull structured fields from forms, PDFs or emails into a record.
How it works
- Starts when
- A new document is uploaded.
- What runs automatically
- A draft record with confidence flags. Missing fields are never guessed.
- Where AI is used
- Extraction only.
- Where a person is in charge
- A person verifies key fields, especially anything clinical.
AI quality assurance
Prove an AI flow is safe enough to run, and stays that way.
How it works
- Starts when
- Before launch, after any change to a prompt, model or data source, and on a schedule.
- What runs automatically
- A test set of at least 30 realistic cases, a test report and a version log.
- Where AI is used
- May help generate test cases, reviewed by a person.
- Where a person is in charge
- A named person signs off on results and approves changes.
No flows match. Try another word, or book a call and we’ll find the right fit.
Build size: S small · M medium · L larger · XL major build, with a written safety protocol. Pricing depends on size, so ask on a call.
Sample library. Every flow is adapted to your voice, your tools and your rules before it is built.
From first call to handed over.
Not sure where to start? Start free.
Try a free starter kit and set up your first flow yourself, or book a call and we’ll map it together.
Prefer email? admin@atuned.org
Scope. Atuned designs and builds systems. It does not provide clinical services, and every flow offers general wellbeing support, not medical or mental health treatment. The practice remains responsible for its professional obligations, clinical decisions and legal compliance, including privacy and health-data rules where it works. If you are in danger or think you may harm yourself, contact your local emergency number (988 in the US). Privacy · Terms