The lighthouse approach to personal AI

Fanar, named after the Arabic word for lighthouse, is building consumer applications that prioritize behavioral adaptation over simple task management. Based in the MENA region and drawing on research from entities like the Qatar Computing Research Institute (QCRI), the company is carving out a niche at the intersection of Arabic-centric large language models and personalized agentic software. While many AI firms focus on horizontal productivity tools for the enterprise, Fanar focuses on the everyday, targeting routines and travel through its primary products, Daylst and Tourida.

Predictive routines with Daylst

Daylst is the company's most developed offering, positioned as an AI routine coach. Unlike traditional habit trackers that require manual input and rigid scheduling, Daylst uses generative AI to build personalized plans based on user-defined goals. The system includes a behavioral learning layer that monitors how users interact with their schedules. It attempts to predict energy patterns and adjust task difficulty to match the user's natural circadian rhythms. This shift from a manual calendar to an autonomous coach represents a move toward agents that manage time on behalf of the user, rather than just recording it.

Arabic-centric foundational technology

The technical backbone of Fanar is more complex than its minimalist consumer site suggests. Research associated with the brand points to the development of Fanar as an Arabic-centric multimodal generative AI platform. This includes specialized implementations like MorphBPE, a tokenizer designed to handle the morphological complexities of the Arabic language more efficiently than Western-centric models. This vertical integration—building both a foundational model and consumer applications—is a strategy intended to provide higher accuracy and relevance for a specific regional and linguistic context.

Expanding travel discovery

Beyond daily planning, Fanar is expanding into the travel sector with Tourida. Currently in development, Tourida is an AI travel companion that moves beyond keyword-based search results. It is designed to match travelers with hostels, hotels, and activities based on individual travel styles and adventure preferences. By focusing on off-the-beaten-path recommendations and camping, the company is attempting to capture a demographic that values discovery over standard tourism packages.

Integration and the Model Context Protocol

A notable development in the Fanar ecosystem is its early engagement with the Model Context Protocol (MCP). Development records indicate the existence of a fanar-mcp-app designed to bridge the gap between the Fanar LLM and modern tool-augmented AI systems. This indicates that the company is building a framework where its models can interact with external data and tools. For users, this creates the potential for a more unified agent experience where a routine coach can communicate with a travel assistant through a shared architectural standard.