Sli — Platform Overview for Coaches

The Problem This Solves

AI is reshaping how people work, learn, parent, teach, and make sense of themselves. A marketing director watches AI generate in 30 seconds what her team produces in a week. A parent can't help their child with homework because the way learning works has changed. A teacher finds their students submitting AI-generated work and has no framework for responding.

These aren't skills gaps. They're identity problems. "I'm a marketing leader." "I'm a good parent." "I'm an expert teacher." AI puts a question mark after every one of these statements.

Governments offer generic reskilling. Companies hand out LinkedIn Learning licences. The tech industry says "learn to prompt." All of these treat a deeply personal crisis as a curriculum problem. A course can teach someone what AI does. It can't help them figure out what they do now that AI does what they used to do.

That requires a coach. Someone who understands their context, their history, their strengths, and their fears.

There are 122,000 coaches globally. The disruption touches billions. The maths doesn't work — unless coaches can scale.

What This Platform Is

Infrastructure that lets you deliver your coaching methodology at scale — without losing the personalisation that makes coaching work.

You upload your content. The platform turns it into structured learning, AI-powered coaching conversations, and a personalised intelligence feed. The AI doesn't have its own methodology. It carries yours. If the AI sounds generic, the platform has failed. If it sounds like you would sound, it's working.

A single coach typically supports 20-30 people deeply. With this platform, you can support 200-500 with the same quality of personalisation, because the platform remembers each participant's journey and keeps the learning alive between sessions.

How It Works: Three Composable Modes

Every person navigating disruption needs three questions answered:

What is happening? (Learn)
What does it mean for me? (Coach)
What just changed, and what should I do now? (Adapt)

Each mode stands alone. You enable whichever combination serves your participants. Together, they form a complete adaptation system — and critically, they share context. What someone learned informs how they're coached. How they're coached informs how updates are personalised to them.

No other tool does this. Coaches currently stitch together an LMS, a coaching tool, and a newsletter — and lose all continuity between them. This platform connects the thread.

Sources — Your Content, One Place

You upload your materials once: frameworks, session plans, research, philosophy. Tag them by category. Assign them to Learn, Coach, or Adapt — or all three. The same content works differently depending on what the participant needs:

For Learn: your materials become structured modules, quizzes, and interactive lessons
For Coach: your philosophy and frameworks become the invisible scaffolding that shapes AI coaching conversations
For Adapt: your content plus external feeds become personalised, contextualised updates

Supported formats: PDF, DOCX, TXT, Markdown, MP3, MP4, WAV, WebM, QuickTime video, PPTX.

Learn — "What is happening?"

You don't teach "Intro to AI." You teach "What AI Means for How You Lead." The content comes from your expertise and source materials, not a generic catalogue.

The platform extracts learning objectives from your uploaded materials and generates structured modules with sessions and quizzes. You review and edit before publishing.

What it includes:

AI-generated modules from your source materials
Four quiz types: multiple choice, true/false, free text (AI-graded), short answer
Per-session chat assistant that knows what the participant is currently studying — when they ask "why is question 2 wrong?", the AI knows what they're referring to
Adaptive recommendations based on performance
Progress tracking: completion rates, quiz scores, objectives mastered
Quick-start pipeline: upload content, extract objectives, generate modules — live in hours

Coach — "What does it mean for me?"

The marketing director who just completed a Learn module on "Leading Through Technological Disruption" and scored low on "Comfort with Ambiguity" doesn't need generic coaching. She needs coaching that knows her learning history and meets her where she is.

The AI coaches in the style and framework you defined. Your source documents inform the system prompt and shape coaching responses. The participant never sees the sources directly — they experience them as the quality and specificity of coaching.

What it includes:

Five coaching modes: open session, scenario coaching, roleplay, framework interpretation, reflection
Eight-layer prompt architecture: you control the base, frameworks, boundaries, mode behaviour, context, and session layers independently
Persistent memory (with participant consent): the AI remembers goals, patterns, and progress across sessions. Memory is opt-in, and opting out triggers immediate hard deletion.
Dual-layer boundary detection: rule-based patterns plus AI classifier. Self-harm detection is always on with immediate escalation to you via email. Categories are configurable.
Session analysis: each coaching session is analysed against your competency framework, producing dimension scores and growth tracking
Emergency disable: if a participant is in crisis, you can immediately suspend their access and close all sessions
All coaching content encrypted at rest

Adapt — "What just changed, and what should I do now?"

This is not a news feed. This is the thing you wish you could provide but can't — ongoing developmental support between sessions, at scale.

When OpenAI ships a new capability, the marketing director doesn't need a news article. She needs someone to tell her: "Remember when we worked on redefining your team's value proposition around strategic thinking? This new capability makes that more important, not less. Here's why." That's what you would say — and Adapt says it at scale, for every participant, personalised to their journey.

You configure external sources — RSS feeds, URLs, YouTube channels, documents. The platform monitors them for new content. When something arrives, it's summarised and then personalised for each participant based on what they've learned and what they've been coached on.

The same article reaches 200 participants. Each one gets a different interpretation.

What it includes:

Source types: RSS feeds, web URLs, YouTube channels, document uploads
Configurable polling frequency, keyword filters, and content age limits
Per-participant personalisation using their Learn and Coach history
Relevance scoring so participants see what matters most to them first
Intelligence synthesis: topic clusters and AI-generated briefings across all sources
Personalised content encrypted at rest

Measuring Impact

Upload your coaching materials and the platform extracts a competency framework — 3 to 8 dimensions with proficiency levels. You review and approve it before it goes live.

From then on, every coaching session is automatically analysed against that framework. You can see not just "how many sessions happened" but "47 people moved from Developing to Proficient on Strategic Thinking this quarter."

What it includes:

AI-extracted frameworks from your uploaded content
Per-participant dimension scores tracked over time
Trend analysis: improving, stable, or declining per dimension
Strengths and growth areas identified automatically
Cohort-level insights: how a group is developing, common themes, score distribution
Development passport: a verifiable credential participants can share showing their growth across all three modes

For enterprise buyers, this is ROI evidence. For you, it's proof that your methodology works.

Cohort Management

Organise participants into groups. Bulk enrol. View cohort-scoped analytics — how the group is progressing collectively, not just individual by individual.

How This Compares to Other Tools

Enterprise coaching platforms (BetterUp, CoachHub)

What they do well: Massive scale. Deep enterprise integration. HR dashboards. Dedicated human coach matching. Proven at Fortune 500 companies.

Where they fall short for you: Coaches are interchangeable labour — the platform owns the participant relationship, the data, and the methodology. If you leave, you lose everything. Pricing starts at $2,000-4,000 per employee per year, sold to HR departments. Independent coaches and small firms are priced out entirely. AI coaching (where it exists) uses the platform's approach, not yours.

How this platform differs: You own everything — methodology, data, participant relationships, brand. You control the AI's behaviour layer by layer. Your methodology is the product, not the platform's. You can export your entire configuration and self-host if you want to leave.

Practice management tools (Simply.Coach, Delenta, CoachAccountable)

What they do well: Scheduling, invoicing, goal tracking, session notes, client management. Mature, coach-friendly interfaces. Affordable ($9-29/month).

Where they fall short for you: No AI coaching capability. No content delivery or learning system. No personalisation. They solve administration — scheduling and billing — but they don't help you scale your methodology to more people. A coach using Delenta still coaches 20-30 people manually.

How this platform differs: This isn't a practice management tool. It doesn't handle scheduling or invoicing. It handles the coaching itself — delivering your methodology via AI to hundreds of participants. You'd likely use both: a practice tool for your 1:1 human practice, and this platform for your scaled programme.

Where they're ahead: Scheduling, invoicing, client management. This platform doesn't do those things. If you need a CRM for your coaching practice, you'll still need a separate tool.

AI white-label coaching (Rocky.ai)

What they do well: Branded AI coaching app. Quick to set up. Daily 5-minute coaching chats. Mobile-first experience. Low price point ($10/user/month).

Where they fall short for you: It's a branded ChatGPT wrapper. The AI doesn't carry your specific methodology — it uses a generic coaching approach with your branding on top. No structured learning. No contextualised adaptation. No competency frameworks. No session analysis. No boundary detection. No encryption of coaching content. No persistent memory across sessions.

How this platform differs: The AI coaches in your methodology, informed by 8 layers of configuration you control. It remembers each participant's journey. It detects when conversations cross boundaries and escalates to you. Coaching content is encrypted. Sessions are analysed against your competency framework.

Where they're ahead: Mobile app experience. Quick onboarding — a coach can be live in minutes. Simpler interface for participants who just want a quick daily check-in rather than a structured programme.

Course platforms (Thinkific, Kajabi, Circle.so, FreshLearn)

What they do well: Mature course builders. Landing pages and marketing funnels. Email sequences. Payment collection from students. Community features (especially Circle). Mobile apps. Video hosting. Certificates. Affiliate programmes. Large integration ecosystems (Thinkific has 100+ apps).

Where they fall short for you: Their AI generates marketing content, not coaching. Thinkific's "Thinker AI" answers questions about course content but doesn't coach. Circle's AI agents are the closest — they can answer FAQs from a knowledge base — but they have no persistent memory, no coaching methodology, no session analysis, and no personalisation per individual. None of them connect learning to coaching to ongoing support. They solve "sell courses online," not "scale my coaching methodology."

How this platform differs: Three modes with shared context instead of siloed courses. AI that coaches in your methodology instead of answering FAQs. Per-participant personalisation instead of one-size-fits-all content. Competency frameworks and session analysis instead of completion percentages. Boundary detection and safety controls built in. Full data sovereignty instead of proprietary content formats.

Where they're ahead — significantly:

Marketing and sales infrastructure. Landing pages, email sequences, checkout flows, affiliate programmes, sales funnels. This platform has none of that. You need a separate tool to market and sell your programme.
Participant payment collection. Course platforms let your students pay you directly. This platform bills you, not the participant.
Community and social features. Circle especially has mature community spaces, live rooms, DMs, events. This platform has messaging (coach↔participant, peer chat within cohorts, session requests) but no forums, live rooms, or events.
Video content. Course platforms have more mature video hosting. This platform supports video and audio upload with AI transcription (Whisper), inline media players with transcript sync, and playback speed controls — but lacks advanced features like adaptive bitrate streaming or video analytics.
Mobile apps. Most offer branded native apps. This platform is responsive web only.
Ecosystem and integrations. Course platforms have years of integration partnerships. This platform has Stripe for billing and Resend for email — that's it.
Market maturity. These platforms have millions of creators, proven business models, established support ecosystems. This platform is early-stage.

The honest summary: if you primarily need to sell self-paced courses online, a course platform is probably the right tool. If you need AI to deliver your coaching methodology at scale with personalisation and continuity across learning, coaching, and ongoing support — that's what this platform does and they don't.

Safety and Compliance

This is a coaching platform, not a therapy tool. Every interaction pathway respects that boundary.

What protects participants

Dual-layer boundary detection: rule-based patterns catch known risks instantly. An AI classifier catches ambiguous cases. Self-harm is always checked first — immediate escalation, never continue coaching.
When self-harm is detected, you receive an email immediately with session context (never message content).
Emergency disable: suspend a participant and close all their sessions with one action.
AI disclosure at every touchpoint: session banner, first message, system prompt, exports, journey dashboard. Participants always know they're interacting with AI.
Consent-based memory: opt-in (default off). Opting out triggers immediate hard deletion. No grey areas.
Ephemeral sessions: messages are deleted when a session closes unless the participant explicitly chose to save them.
Full data control: participants can view, export (JSON, CSV, PDF), or delete all their data at any time. Deletion is permanent — no soft deletes, no recoverable data.
All coaching conversations encrypted at rest (AES-256-GCM). The database stores ciphertext, not plaintext.
PII-safe logging: emails, IPs, and tokens are stripped from all error logs.

Regulatory posture

GDPR: hard-delete erasure, data portability in machine-readable formats, encryption by design, consent-based processing, automated retention cleanup, audit trails that never log content. All critical and high-priority compliance issues identified and resolved.
EU AI Act: limited-risk classification for coaching. AI transparency at 7+ touchpoints. Boundary detection as risk management. Human oversight via admin dashboard and escalation. The EU AI Act high-risk enforcement deadline is August 2, 2026 — this platform is ready.

Data isolation

Row-Level Security on every tenant table — one coach's data is invisible to another's
Every database query is scoped by tenant ID and user ID
LLM providers are swappable — no vendor lock-in on where your data is processed

For coaches operating in Europe, compliance isn't optional. For coaches anywhere, it's a trust signal to enterprise clients. Most competitors claim GDPR compliance but none address the EU AI Act. None mention encryption by design or hard-delete semantics.

Admin Dashboard

This is not a developer tool. It's a product interface for running your coaching practice.

User management: invite, suspend, remove. Activity indicators show who's active, online, or inactive.
Source management: upload, tag, assign to modes.
Coaching configuration: edit prompts layer by layer, set boundaries, configure modes.
Learning management: review generated modules, publish, track enrolments.
Adapt pipeline: source health, ingestion logs, personalisation settings.
Competency frameworks: review, edit, approve AI-extracted frameworks.
Safety dashboard: boundary trigger frequency, escalation events.
Analytics: usage metrics, token costs, engagement trends.
Cohort management: create groups, bulk enrol, cohort-scoped analytics.
Platform settings: branding (logo, colours, names), LLM provider selection, retention policies.
Full config export/import: your entire setup is portable.
Setup wizard: guided first-run. Upload content, platform builds your first programme. Not a settings dump.

White-label by default. Your logo, your colours, your name. No extra cost.

Embeddable chat widget for existing websites or member portals. Same coaching quality, same boundary rules.

Pricing (TBD)

The platform charges a subscription to the coach — not per participant. The pricing model, tiers, and thresholds are not finalised yet.

How you price your programme to your own clients is entirely up to you. Free, tiered, per-cohort, employer-funded — your business model, your decision. The platform doesn't take a cut and doesn't impose a pricing structure on your participants.

What It Doesn't Have Yet

Messaging Phase 2 (Phase 1 delivered)

Phase 1 is live: coach-to-participant, participant-to-coach, and peer-to-peer messaging within cohorts. Encrypted at rest (AES-256-GCM). Session requests let participants ask for a human coaching session. Admin messaging page for coach-initiated conversations. Peer chat toggle per tenant.

What's planned for Phase 2:

Coach broadcasts to everyone, a cohort, or an individual — with optional AI personalisation that contextualises your message to each participant's journey, without you seeing their private coaching data.
AI facilitator in group conversations — surfaces themes, asks questions, draws out discussion using your methodology. Never references private coaching sessions.
Transport layer: email, WhatsApp, or other channels beyond in-platform.
Push/in-app notifications for new messages.
Typing indicators and online presence in channels.
File/media attachments in messages.

Voice coaching

Video and audio content is supported in learning sessions — upload, Whisper transcription, inline players with transcript sync, playback speed controls. Text-to-speech is available for learning content. However, voice coaching conversations (real-time speech-to-text/TTS in Coach mode) are not built. AI video and avatar coaching are further out — latency and emotional realism remain unsolved at the quality level this requires.

Marketing and sales infrastructure

No public programme pages, no landing page builder, no email marketing, no sales funnels, no affiliate programme. You need external tools for marketing and selling your programme.

Participant payments

The platform bills you, not the participant. No mechanism for participants to pay you through the platform.

Community

Peer-to-peer messaging exists within cohorts (Phase 1). No forums, live rooms, or events beyond messaging channels.

Mobile app

Responsive web. Works on mobile browsers. No native app, no push notifications.

Scheduling and practice management

No session scheduling, no invoicing, no CRM. This handles the coaching delivery, not the coaching business administration.

Certificates and SCORM

No per-module completion certificates. No LMS integration. The development passport covers cross-mode progression but individual module certificates aren't generated.

Readiness scoring

Aggregated Adapt engagement metrics showing whether participants are keeping up with changes in their field. Planned, not built.

What You Own

Everything.

Your methodology — frameworks, prompts, source materials. Exportable at any time in standard formats.
Your configuration — full export/import. Take it to another platform or self-host via Docker.
Your participant relationships — no lock-in, no proprietary formats.
Your brand — white-label is the default.
Your LLM choice — swap between OpenAI, Anthropic, or Google via configuration. No code changes.

Enterprise platforms own the platform and the data. If a coach leaves, they lose everything. This platform mandates sovereignty. For coaches who've spent decades building methodologies, that isn't a feature. It's a trust requirement.

Who This Is For

You are	You'd use	Why
Leadership coach running AI transition programmes	Learn + Coach + Adapt	Professionals whose identities are in flux need all three
Educator helping students prepare for an AI world	Learn + Coach	Structured content plus guided practice
Parenting coach helping families navigate AI	Coach + Adapt	Personal guidance plus ongoing updates
Boutique L&D firm serving corporate clients	Learn + Coach + Adapt	Scale without hiring. Compliance for enterprise contracts.
Professional association supporting members	Learn + Adapt	Learning plus contextualised updates across the membership
Career transition specialist	Coach + Adapt	Application plus ongoing support as the landscape shifts
Community educator working with older adults	Learn + Coach	Understanding what's changing plus guidance navigating it

The Bigger Picture

122,000 coaches can serve about 3 million people at current capacity. With infrastructure like this, they could serve 25-60 million — without diluting the expertise that makes coaching transformative.

The AI transition isn't a workforce problem. It's a human adaptation problem. The people best equipped to help — coaches, educators, guides — need tools that match the scale of the challenge.

This platform is that tool.