Product guide

Everything you can ask

Avsar is one research chat. You ask in plain English; the agent writes and runs the analysis and shows you the result. Save the good ideas as strategies and paper trade them with virtual capital.

How it works

1Ask

Pose a question in the research chat

2Get analysis

Charts, tables and backtests as artifacts

3Save

Keep the good ones as strategies

4Paper trade

Forward-test with virtual capital

Ask in plain English

in chat

One chat. The agent writes and runs real analysis, then shows you the result.

1
Type a question
Ask the way you'd ask an analyst — "show me a chart of RELIANCE over the last year". No menus, no query builder.
2
The agent does the work
It writes Python, runs it against years of market data, and renders charts, tables and numbers as artifacts in the thread.
3
Pick your universe
Switch between NIFTY100, NIFTY50 and SENSEX in the composer to scope what a question runs over.
4
Iterate in follow-ups
"Now add a 50-day moving average" or "compare it to INFY" — the thread keeps context so each step builds on the last.

Try these prompts

Show me a chart of RELIANCE over the last yearTry it

What's RELIANCE's PE and 52-week high?Try it

Screen the market

in chat

Filter the universe by any rule — fundamentals, momentum, technicals — in one sentence.

1
Describe the filter
Combine fundamentals and price action freely: valuation, returns, ROE, moving-average position, and more.
2
Get a ranked table
Matches come back as a sortable table you can read at a glance — no spreadsheet export needed.
3
Refine without restarting
"Now only the ones above their 200-day average" narrows the same result set in the next message.

Try these prompts

High ROE, low PE quality stocks in NIFTY50Try it

Stocks that went up more than 20% in the last 10 daysTry it

Backtest an idea

in chat

Describe a strategy; get an equity curve and the metrics that matter.

1
State the rules
Entry and exit in plain language — a moving-average cross, an RSI threshold, a time-based exit.
2
Run over real history
The agent codes and runs it across years of daily (and minute) data, with realistic costs.
3
Read the scorecard
Equity curve plus Sharpe, win rate, max drawdown and trade count — the numbers you'd actually judge it on.
4
Tune it
"Add a 2% stop-loss" or "try the 20/100 cross instead" re-runs with the change in the same thread.

Try these prompts

Backtest a 9/50 MA cross on INFY for the last 2 yearsTry it

Analyse options

in chat

Option chains with Greeks, and payoff diagrams for multi-leg ideas.

1
Pull a chain
Ask for an index option chain to see strikes, IV, open interest and the full Greeks.
2
Sketch a structure
Describe a spread or straddle and the agent plots the payoff — max profit, max loss, breakevens.
3
Test it on history
Option strategies can be backtested across past expiries for NIFTY, BANKNIFTY, SENSEX, FINNIFTY and MIDCPNIFTY.

Try these prompts

Show the NIFTY option chain for this week's expiry with GreeksTry it

Plot the payoff of a NIFTY iron condor 300 points wideTry it

Dig into a company

in chat

Fundamentals, financial statements, shareholding and peers — across the NSE universe.

1
Ask about the business
Valuation, profitability and growth come back in a clean, comparable view.
2
Open the statements
P&L, balance sheet and cash flow, plus shareholding trends across promoters, FII and DII.
3
Compare peers
Put two names side by side and the agent lines up the metrics that distinguish them.

Try these prompts

Compare TCS vs Infosys fundamentalsTry it

Show HDFC Bank's shareholding trend over the last 8 quartersTry it

Save & manage strategies

/strategies

Turn a good backtest into a saved strategy with code and parameters you can revisit.

1
Save from chat
When a backtest looks right, ask the agent to save it. It becomes a named strategy with its code and parameters.
2
Find it on Strategies
The Strategies page lists everything you've saved, plus a Popular tab of curated examples to start from.
3
Iterate over time
Reopen a strategy and ask for changes — each edit is captured so you can compare versions.

Try these prompts

Save this as a strategy called 'INFY MA cross'Try it

Paper trade it

/paper

Deploy a strategy with virtual capital and track how it performs against the live market.

1
Deploy to paper
Push a saved strategy to paper trading — no real money, no broker. It runs on scheduled checkpoints against real market data.
2
Track performance
The Paper page shows each deployment's NAV, today's P&L and return, with a decision log of what it did and why.
3
Forward-test with confidence
Watch a strategy behave in real conditions before you'd ever consider trading it for real.

Try these prompts

Deploy this strategy to paper and track its daily P&LTry it

What's under the hood

in chat

The data the agent can see — so you know what it can answer before you ask.

1
Equities
~10 years of daily candles and ~5 years of 1-minute candles across ~1,000 NSE stocks, with Nifty 50 / 100 / 250 / 500 / 750 cap-rank tiers ready to screen.
2
Options
Full historical option chains for NIFTY, BANKNIFTY, SENSEX, FINNIFTY and MIDCPNIFTY, plus F&O stocks — strikes, expiries and Greeks.
3
Fundamentals
Financial statements, ratios, shareholding and peers across the NSE universe, sourced from public filings.

Still stuck? Just ask.

The agent can describe its own capabilities — “what can you do?” works too.

Open research