Mikav: An Open-Source AI Copilot Bridging Cultural Heritage and Creative Entrepreneurship in Kerala
Presents the full Mikav system architecture (dataset → model → copilot → dev platform), deployment approach, and a case study from the SparkX cohort showing real-world usage and outcomes.
Abstract
We present Mikav, an open-source AI copilot that bridges cultural heritage preservation and creative entrepreneurship in Kerala. The system integrates a curated Malayalam culture corpus, fine-tuned language models, an interactive copilot interface, and a startup bridge platform. We describe the full architecture (dataset → model → copilot → developer platform), our deployment approach, and present a case study from the SparkX incubation cohort demonstrating real-world usage and outcomes.
1. Introduction
Kerala possesses extraordinary cultural wealth — from performing arts like Kathakali and Theyyam to literary traditions, festivals, and craft heritage. Yet this knowledge remains difficult to access digitally, particularly in Malayalam. Simultaneously, a growing ecosystem of culture-tech entrepreneurs seeks to build businesses around heritage. Mikav connects these needs.
2. System Architecture
2.1 Data Layer
- Open Malayalam culture corpus
- Community-contributed datasets
- Verified knowledge graph
2.2 Model Layer
- Fine-tuned Malayalam LLM
- Cultural knowledge embeddings
- Human-in-the-loop verification
2.3 Application Layer
- Chat copilot interface
- Culture map explorer
- Open API
2.4 Platform Layer
- Developer tools and SDKs
- Startup bridge features
- Community governance
3. Design Principles
- Open by default
- Malayalam-native, not translated
- Community-verified
- Entrepreneur-friendly