YIMO — Your Intelligent Motive Optimiser
Inspiration
High-quality coaching changes lives.
The right question at the right time can unlock clarity, confidence, and forward movement.
But coaching is often expensive, time-intensive, and limited to a small group of already-optimised individuals.
At the same time, most AI tools provide answers — not perspective.
YIMO was inspired by a simple question:
What if AI coaching felt calm, structured, and psychologically grounded — without requiring users to build complex systems or learn technical tools?
Instead of automation for its own sake, the goal was intentional thinking made accessible.
What it does
YIMO is a structured AI coaching environment designed around how humans actually grow.
At its foundation are five system coaches, aligned with Maslow’s hierarchy of needs:
- Stability → Max
- Security → Atlas
- Connection → Luna
- Growth → Ace
- Self-Actualisation → Zen
In Ask Mode, every question is routed through the most relevant psychological lens. Responses are never generic — they come from a defined coaching perspective.
Beyond Ask Mode, YIMO offers a growing library of specialised coaches focused on specific domains like career, productivity, relationships, finance, and long-term strategy.
Users can also:
- Create their own personalised coach
- Customise its tone, style, and behavioural approach
- Assign it a unique voice, giving it a distinct personal dimension
Each coach has its own persona and communication style, making interactions feel intentional rather than uniform.
Coach-specific chats are deeply personalised. Conversations evolve over time and adapt to the user’s goals, patterns, and context.
YIMO operates with two layers of memory:
- A central memory layer where users can add notes, documents, and links once
- A coach-specific memory layer, where each coach maintains its own evolving understanding
This creates both shared context and specialised depth.
YIMO can also automatically retain meaningful insights when they emerge, strengthening long-term continuity without requiring manual effort.
The ecosystem is expandable:
- Users can share their custom coaches with others
- They can publish coaches to the community
- The coach library grows organically through contribution
This keeps the coaching ecosystem dynamic and community-driven.
YIMO integrates with tools like Google Calendar and Notion so that reflection can translate into scheduled action and real-world execution.
In simple terms:
[ Better\ Progress = f(Context,\ Structured\ Perspective,\ Continuity) ]
Where:
- Context comes from central and coach-specific memory
- Structured Perspective comes from psychologically aligned coaching lenses
- Continuity builds long-term clarity
How we built it
YIMO was built as a modular coaching architecture with clear separation between experience and intelligence.
Core components include:
- A psychological routing engine that selects the appropriate need-based system coach in Ask Mode
- A flexible coach framework supporting system, library, and user-created coaches
- Dual memory layers — a shared central memory and independent coach-specific memory
- A persona layer allowing tone, behavioural style, and custom voice configuration
- Community publishing infrastructure for sharing and discovering coaches
- Secure integrations with tools like Google Calendar and Notion
The focus was simple on the surface, structured underneath.
Challenges we ran into
- Ensuring responses felt grounded in defined coaching perspectives rather than generic AI output
- Designing dual memory systems without creating cognitive friction
- Balancing deep customisation with usability
- Maintaining coherence across community-published coaches
- Preserving a calm user experience while supporting architectural complexity
Accomplishments that we're proud of
- Creating a psychologically structured, non-generic coaching system
- Implementing dual-layer memory for depth and continuity with multimedia support
- Enabling fully customisable coach personas with distinct voices and chat style
- Launching a shareable, community-driven coach ecosystem
- Scheduled actions to automatically trigger user requests daily
What we learned
Structured response routing produces more consistent coaching outcomes than unconstrained generative output.
Layered memory architecture (central + coach-specific) significantly improves contextual coherence and longitudinal interaction quality.
Persona configuration (tone, behavioural constraints, and voice modulation) materially affects user perception of depth, authority, and alignment.
Explicit psychological scaffolding reduces randomness in responses and improves decision-support reliability.
Simplicity at the interface level requires strict separation between orchestration logic and presentation layer.
What's next for YIMO
- Expanding the coach ecosystem and discovery system with 999+ coaches covering every domain
- Improving routing precision across psychological dimensions
- Enhancing memory intelligence and pattern awareness
- Introducing advanced coaching modules
- Deepening third-party integrations
The long-term direction is to evolve YIMO into a scalable, psychology-driven coaching ecosystem that remains calm, accessible, and continuously adaptive.
Built With
- dart
- docker
- fastapi
- firebase-(auth-&-messaging)
- flutter
- google-cloud-run
- google-gemini-api
- langgraph
- openai-api-(gpt-4o)
- openai-embeddings
- pgvector
- postgresql
- revenuecat
Log in or sign up for Devpost to join the conversation.