Monafy is a research platform for Turkish and US stock markets, bringing financial data, company disclosures and news into one product. The aim is to keep the information needed to examine a company or a market session connected throughout the research process.
The product works in Turkish and English. A reader can move from a market overview to a company, inspect a source disclosure and ask the AI assistant for help without treating each as an unrelated task. The tools answer different questions while preserving the context of the research.
Understanding a market session
The Market Report examines price moves and market participation alongside news, company disclosures and the economic calendar. Pre-market, live-session and closing reports need different context. A previous session's results should not quietly become a description of the current trading day.
The figures come from the financial data pipeline. AI writes explanations using that information; it does not establish the prices or sector performance. Describing a relationship between a news event and a market move also needs language that reflects the available evidence.
Connecting the parts of company research
Company pages bring prices, financial statements and disclosures into the same research workflow. Turkish company announcements and US filings can be inspected with links to the original material. An AI summary helps a reader locate the relevant point without replacing the source document.
The screener helps identify companies matching selected characteristics. Watchlists preserve the next step in the research, while price alerts track a chosen condition. Current availability and usage limits are described on the features and pricing pages rather than fixed into this article.
Giving the assistant access to product data
The Monafy Assistant is connected to the application's financial tools. Depending on the question, it can read relevant stock information, disclosures or watchlist data. It is intended to use the product's information rather than rely entirely on general model knowledge.
The model chooses tools and writes an explanation, while the interface receives data blocks through a defined schema. Free-form language and numerical presentation therefore have distinct responsibilities. Proposing a user action is also separate from executing that action in the application.
Building this experience requires service design as well as prompt engineering. The available tools, the context supplied to the model and the response format all shape the behavior a user encounters. An effective prompt depends on those surrounding decisions.
Keeping time and coverage visible
A price needs a meaningful timestamp and an understandable delay. Upstream publication delay and the processing time added by our own services are different measurements. I discuss the latter in the article on price-stream latency.
A missing value is also different from zero. Sources, freshness and the availability of data matter as much as the explanation built around them. The methodology page is the place to check the current data scope and the product's approach.
Connecting product decisions with engineering
I design product flows and service responsibilities, and direct development with AI agents. Financial data infrastructure, report generation and assistant integration are connected parts of that work. The case studies on this site examine individual engineering decisions within that larger product, so readers can see both the reasoning and its practical limits.