Personal conversational AI
How to use
Pi.
Reflective conversation, brainstorming, goal clarification, rehearsal, and low-stakes personal organization in a warm conversational format.
Pi is a personal conversational AI assistant from Inflection AI. Reflective conversation, brainstorming, goal clarification, rehearsal, and low-stakes personal organization in a warm conversational format. This guide covers the whole path in one place: official access, a first session that produces something reviewable, the checks that make output trustworthy, and the permissions worth limiting before you connect real work.
AI should make the work easier to inspect. If the workflow removes the source, the owner, or the review step, redesign the workflow.
Access & setup
Find, install, and sign in to Pi
Get into the official Pi experience with the right account and a setup you understand.
- 01
Start at https://pi.ai/ and confirm the domain before entering account or payment information.
- 02
Availability: Pi is available through its web experience and supported iOS and Android apps, with text conversation, voice-oriented interaction, limited web search, and synchronized history for signed-in adults.
- 03
A supported browser or mobile device, an account for persistent history, age 18 or older, and an explicit decision about personal information, history retention, model improvement, and notifications.
- 04
Sign in with the account you intend to keep using, then review plan, data, notification, and permission settings.
- 05
Run one low-risk test task before connecting sensitive files, repositories, or workspace data.
- Use official download pages.
- Review permissions during setup.
- Keep installers and applications updated.
First session
Your first useful Pi session
Learn the interaction loop using a small task with a clear outcome.
- 01
Review Pi's current privacy policy, model-training notice, safety limits, and history controls before discussing personal matters.
- 02
Define a low-stakes outcome such as clarifying options, rehearsing a conversation, or turning thoughts into a short plan.
- 03
Share the minimum context needed, distinguish facts from feelings and assumptions, and ask Pi to summarize rather than decide for you.
- 04
Check factual claims and professional implications elsewhere, export useful notes if needed, and delete messages or the account when appropriate.
- State the outcome before the background.
- Provide the real source material.
- Review the result before expanding the task.
Quality control
Check the quality of Pi output
Turn a plausible draft into work you are willing to stand behind.
- 01
Restate the original outcome and acceptance criteria before reviewing the response.
- 02
Mark every factual claim, calculation, quotation, or figure that could materially affect the result.
- 03
Check those items against the original source, a primary reference, or direct inspection.
- 04
Ask the tool to list its assumptions and uncertainty, but never rely on self-review as the only check.
- 05
Complete a human edit for relevance, tone, privacy, accessibility, and the exact delivery format.
- Verify consequential claims independently.
- Fluency is not accuracy; confident phrasing is not evidence.
- Keep a record of sources and review decisions.
Privacy & permissions
Use Pi safely
Use the platform productively without giving it unnecessary access or authority.
- 01
Classify the task and its data before uploading or connecting anything.
- 02
Review the product, account, organisation, retention, and training settings that apply to you.
- 03
Grant the narrowest file, workspace, microphone, screen, or integration permissions needed.
- 04
Remove secrets and personal or regulated information unless your approved policy explicitly permits it.
- 05
Review the output and revoke permissions or disconnect sources that are no longer needed.
- Pi is designed to feel personal and emotionally intelligent, which can encourage over-trust or over-disclosure. It is not a therapist, emergency service, lawyer, doctor, or financial adviser. Inflection states that information may be used to develop and train its models, applies moderation and human review in some cases, and has specific retention rules. Minimize sensitive data, preserve real human support, verify consequential claims, and use history export and deletion controls deliberately.
- Follow your organisation's approved-use policy.
- Never treat fluent output as authorization to act.
Core workflows
Step-by-step ways to use Pi for the work it does best.
Each workflow below has its own outcome, steps and review checkpoints.
Use reflective conversation to clarify priorities without delegating judgment or substituting AI for qualified human support.
↓ 02 WorkflowRehearse an important conversation with PiPractice wording and responses while preserving the other person's privacy and preparing for a real human exchange.
↓Workflow 01
Turn a difficult decision into a grounded plan with Pi
Use reflective conversation to clarify priorities without delegating judgment or substituting AI for qualified human support.
Outcome: Leave with a short, reversible next-step plan that distinguishes feelings, assumptions, facts, and decisions.
- 01
State the low-stakes decision, deadline, options, constraints, and what kind of support you want while omitting identifying or highly sensitive details.
- 02
Ask Pi to reflect your goals and concerns, then correct any mistaken assumptions before requesting suggestions.
- 03
Generate several options with benefits, costs, uncertainties, reversible first steps, and questions that require a person or authoritative source.
- 04
Check factual, health, legal, financial, workplace, and relationship implications with appropriate independent sources or trusted people.
- 05
Choose one small action yourself, schedule a review point, export only useful non-sensitive notes, and delete conversation material you do not want retained.
What good use looks like
- Use Pi for reflection, not authority.
- Keep real people in the support loop.
- Prefer reversible next steps.
Workflow 02
Rehearse an important conversation with Pi
Practice wording and responses while preserving the other person's privacy and preparing for a real human exchange.
Outcome: Create a respectful opening, listening plan, and boundary for a conversation you will conduct yourself.
- 01
Describe the purpose, relationship, desired outcome, tone, and safety constraints using roles instead of names or identifying details.
- 02
Draft a concise opening built from observable facts, your own feelings and needs, and a genuine question rather than a diagnosis or accusation.
- 03
Role-play several plausible responses, including disagreement and silence, while asking Pi to flag manipulative, coercive, or escalating language.
- 04
Revise for your authentic voice, prepare active-listening phrases, boundaries, a pause option, and real-world support if the conversation feels unsafe.
- 05
Conduct the conversation without reading a generated script mechanically, respect the other person's agency, and reflect afterward with minimal private detail.
What good use looks like
- Anonymize the other person.
- Do not use AI to diagnose motives.
- Prioritize real-world safety and consent.
Official references
Check the current product documentation.
Features, plan limits, availability, and data controls change. These official pages are the starting points used for this guide.