Built an AI-powered emotional wellness platform for daily emotional clarity

A US-based emotional wellness platform designed to help users name their emotions, identify reactive patterns, and align daily actions with their core values. Built for iOS and Android, it delivers personalised support through daily check-ins, guided micro-audio exercises, journaling, and AI-driven habit coaching.

Unlike generic wellness apps, the platform integrates AI personalisation from day one — adapting every user's experience to their emotional profile, attachment style, and schema roles.

• No platform existed to deliver personalised emotional coaching at scale across a daily user routine
• AI personalisation needed to be embedded from the start — not added as a later feature
• Complex onboarding logic required dynamic audience segmentation, attachment style quizzes, and schema role mapping
• Admin team needed a non-technical portal to manage all content without developer dependency
• Offline capability was essential for core workflows — check-ins, journaling, and reflections — with seamless sync on reconnect

1. Embedded AI from day one

AI personalisation wasn't scoped as a later phase. We designed the entire platform around it from the start — emotion classification, segment-based content delivery, journaling summarisation, and context-aware nudges all built into the core architecture.

2. Built onboarding as the personalisation foundation

The onboarding flow — audience segment selection, attachment style quiz, and schema role mapping — was designed to establish a personalised user profile immediately, so every subsequent experience felt relevant from the first session.

3. Designed for non-technical content ownership

The admin portal was built so the client's team could manage all content — emotions, prompts, audio library, quizzes, and micro-actions — without raising a development request. Content autonomy was a non-negotiable design constraint.

4. Delivered across a structured four-month programme

We ran the engagement in two-week sprints across business analysis, UI/UX, mobile and backend development, AI integration, and UAT — with sprint-end demos throughout to validate features before moving forward.

Mobile Application (iOS & Android)

• Audience segment selection with AI-powered LLM classification at onboarding
• Attachment Style Quiz (15 questions) with GPT-generated personalised result explanations
• Schema Role Quiz (10 questions, multi-role) with adaptive LangChain conversational flow
• Daily emotional check-in with NLP emotion classification and free-text input support
• Reflection and alignment journaling with speech-to-text and AI-powered summarisation
• Guided micro-audio streaming from S3 with tier-based access control
• AI-generated micro-actions tailored to current emotion, schema role, and values
• Dream analysis module with AI interpretation
• Free and premium subscription tiers with in-app billing via StoreKit and Google Play

Admin Web Portal

• CRUD management for segments, quizzes, emotions, prompts, micro-actions, and audio library
• Personalisation rules engine mapping emotions, segments, and roles to content
• Analytics dashboards for engagement trends, emotional patterns, and subscription reporting
• User management with profile view, account controls, and GDPR compliance workflows
• Role-based access: Admin, Content Manager, and Analyst roles

AI Integration Layer

• LLM orchestration via LangChain with server-side prompting and guardrails
• Embedding service using OpenAI or SentenceTransformers stored in pgvector
• Semantic retrieval matching user state to relevant prompts, actions, and audio
• Context-aware nudge scheduling via Redis and BullMQ
• Monthly AI-generated trend reviews and habit coaching summaries per user

Infrastructure & Security

• NestJS monolith with microservices-ready architecture on Docker
• PostgreSQL with pgvector for semantic search and embeddings
• GDPR-compliant data management with export, delete, and consent workflows
• TLS encryption, JWT with refresh rotation, and least-privilege RBAC throughout

Classic Informatics delivered a complete AI-powered emotional wellness ecosystem — mobile app, admin portal, and AI integration layer — designed to give every user a personalised emotional coaching experience from their first session, at scale.

Business Impact

• AI personalisation embedded from day one across onboarding, check-ins, and content delivery
• Admin team can manage all content independently without developer involvement
• Offline-capable core workflows ensure uninterrupted user experience in any connectivity condition
• Scalable architecture designed to grow the user base without re-platforming
• Platform positioned for continuous AI improvement as user data accumulates over time