Why tradecli exists
Indian retail traders have world-class brokers but fragmented tooling. tradecli is one terminal-based AI companion that learns with you, researches for you, and acts through the platforms you already use.
Understanding the system
Most traders are forced to react to fragments: a chart in one tab, broker data in another, news somewhere else, and more signals arriving than any one person can manually process.
tradecli is built around a different view: markets are part of a living global system. What happens in currencies, commodities, rates, earnings, policy, liquidity, and institutional flows all leaves traces. The job is to connect those traces before they become obvious on the chart.
Our vision is an AI trading environment that studies the world like a simulation: every tick, signal, regime shift, and capital movement quantified into context, direction, and probabilistic market expectations.
That same intelligence should also change how a trader works. Modern traders face price action, news, filings, macro data, liquidity, options activity, broker views, and watchlists all at once. tradecli automates the research, monitoring, browser workflows, and execution context around those signals so traders can spend less time juggling tools and more time deciding what matters.
Three layers
We're building this in three layers.
Layer 1 — Application (today). A terminal-first AI agent that drives the browser, automates trader workflows, and helps connect research, signals, notes, and execution context across the tools traders already use. The first job is reducing the manual work of tracking too many moving inputs.
Layer 2 — Market engine. A backend layer for rules, market state, signal history, and simulation. This is where tradecli starts turning scattered inputs into a structured view of regimes, flows, and market behavior over time.
Layer 3 — Trading agents. Specialized agentic models trained for financial workflows: reading context, tracking many signals at once, reasoning across assets, testing scenarios, and helping traders form probabilistic expectations instead of reacting to isolated alerts.
On data: we don't own market data, and aggregation is genuinely hard. So we lean on the browser — broker platforms, Screener, TradingView — as the source of truth today. Over time, the engine becomes the layer that remembers, structures, and learns from what the trader sees.
Who's building this
tradecli is built by a two-person team working on it alongside full-time jobs. That keeps us close to the problem: every feature has to earn its place, every workflow has to save real time, and every release has to move the product closer to the trading assistant we want to use ourselves.
We're keeping the team page lightweight for now while we focus on shipping. If you want to get in touch: contact@tradecli.in.