Discover portfolio strategies with AI agents.
Describe your investment idea and how you want your portfolio to work. Mirtelo’s AI agents search our U.S. market data, build and backtest strategy variations, and return a shortlist with clear portfolio rules, returns and drawdowns.
You approve the research plan.
The agents do the research.
70 TB of historical trades and quotes processed.
Options data, derived Greeks and proprietary features across fundamentals, prices and macro.
Your idea. Our data.
AI-driven discovery.
Get portfolio strategies built around your thesis, with clear rules, full backtests and the evidence to decide what belongs in your portfolio.
Give your idea direction
Tell Mirtelo what you believe, what to focus on and how you want your portfolio to work.
Let the agents explore
You approve the research scope. Our agents discover signals and backtest portfolio rules.
Get strategies with evidence
Compare a shortlist with clear rules, full equity curves, returns and drawdowns.
A deep U.S. dataset.
Built for strategy research.
70 TB of historical trades and quotes processed. Options data and derived Greeks, plus proprietary features across options, equity fundamentals, prices and macro — ready for strategy discovery and backtesting.
History shows how rules behaved in the past, not how they will perform.
Equities & market history
Prices, volume, dividends and corporate actions put stock behavior in context.
Company fundamentals
Growth, cash flow, profitability and valuation help examine business quality and price.
ETFs & market flows
Sector, industry and cross-asset relationships reveal where market activity is shifting.
Macro, rates & credit
Economic conditions, rates and liquidity help test how a thesis behaves across environments.
Trades, quotes & microstructure
Spreads, liquidity, order imbalance and trading-flow features add detail beneath daily prices.
Options data & derived Greeks
Options data, derived Greeks and proprietary features reveal signals beyond equity prices.
| NYSE, Nasdaq & NMS | Trading activity, prices and quotes |
|---|---|
| Issuers, SEC filings & fund sponsors | Fundamentals, earnings, filings and fund information |
| Brokerage analysts & publishers | Ratings, estimates and market context |
| FRED, U.S. Treasury & Cboe | Economic data, rates and volatility indexes |
What is a strategy?
Clear rules for how your portfolio evolves.
A strategy is a repeatable set of backtested rules shaped by your investment intent. Its signals and scoring method determine how stocks are ranked, selected and managed within risk limits.
Refresh the data
Update measures of company quality, value and risk as new market data arrives.
Apply its ranking rules
Each strategy has its own scoring rules. This example weights quality and value 50/50 to rank eligible stocks.
Update the portfolio
Hold ten equally weighted stocks, replacing or resizing positions when the rules call for it while position limits and risk controls stay in force.
Review monthly. Refresh the data, re-rank stocks and review portfolio changes each month, applying the same rules within the strategy’s risk controls.
A simple example
“Every month, rank profitable U.S. companies by a 50/50 blend of quality and value. Hold ten equally weighted stocks. Replace those that no longer qualify, keeping exposure within the strategy’s limits and risk controls.”
Rules tested together. Results you can inspect.
Mirtelo tests ranking, allocation and risk rules together across historical periods. It compares candidate results against the research criteria you specify and explains their differences, assumptions and limitations.
See it in the demoYour market brief.
Your candidate strategies.
Discover a strategy, inspect its backtest, then choose how to follow it — yourself or with your AI.
About this demo & data
This interactive demonstration uses fixed plans and sample performance data, not actual strategy returns. Timing is compressed. Plugin setup, account connections, orders, fills and monitoring are simulated; nothing is installed, connected or scheduled. No research job or trade runs here.
Discover. Understand.
Put your strategy to work.
Describe
Describe your thesis, portfolio preferences and risk priorities in chat.
Discover
Review the plan and time estimate. Approve the run; agents discover and backtest the strategies.
Compare
Inspect the shortlist, portfolio rules, full equity curves, returns and drawdowns.
Follow
Follow the rules yourself, or optionally connect your AI to Mirtelo and your brokerage for updates and orders you approve.
Our research engine.
Your next strategy.
We built this research engine to develop strategies for managing our own capital. Our research tools give self-directed investors the same foundation for discovering and testing their own strategies.
How the data and strategy building blocks fit together
Data Algebra prepares historical evidence: selecting observations, aligning dates and calculating features. Strategy Algebra assembles rules for selection, holding, replacement and allocation. The agent coordinates these steps and explains the outcome. Discovery explores a defined family of combinations; the approved plan sets the scope.
| Quality data, ready for research | Connected market history and company data give each discovery a consistent evidence base. |
|---|---|
| Governed AI, directed by you | Your thesis sets the direction. An approved plan sets the scope. Agents explore the strategy variations. |
| Research you can inspect | See the rules, assumptions and backtests behind each candidate — and understand why it was selected. |
We’re democratizing equity strategy building with governed AI and quality data —
so more investors can turn their ideas into strategies they understand.
Kirill Chufarov · Founder
Data Architect · Chief Investment Officer
Questions, answered
Before you start.
Your idea sets the direction. You stay in control of the strategy and how you follow it.
Who is Mirtelo for?
Self-directed investors who want to turn an investment thesis into a systematic portfolio. Mirtelo brings together market data, strategy discovery and backtesting, with your thesis and constraints guiding the research.
Do I need a fully formed strategy?
An investment thesis or set of portfolio preferences is enough. Describe the opportunity you see, your investment universe and any holding or risk constraints. The agents translate that direction into a research plan and testable strategy variations.
How does discovery work?
Within the research plan you approve, agents examine market data, build variations of portfolio rules and backtest them. Results compare the candidates against the research criteria you specify, with supporting statistics, assumptions and tradeoffs.
What do I approve, and how long does it take?
You approve the research scope: the investment universe, data, strategy variations, constraints and testing approach. The plan includes a runtime estimate based on that scope. You can revise it before the run and track progress during discovery.
What do I actually get?
Market analysis, candidate strategy rules and backtest results. Each candidate includes an explanation of its inputs, selection and ranking approach, portfolio weights, replacement and exit rules, and risk controls. Testing assumptions and performance statistics help you inspect how the rules behaved historically and compare the results.
Is a strategy a fixed list of stocks?
Each strategy has its own combination of signals, stock-ranking rules, portfolio weights and risk controls, discovered around your original request. Those rules are backtested together to show how they behaved historically against your research criteria. The rules guide recurring decisions; the holdings change as new data affects the rankings and risk limits.
How often does a strategy update?
The review frequency is part of the strategy and its backtest — for example, weekly, monthly or quarterly. At each review, the rules are applied to current data to produce a target portfolio. Holdings change only when the selection, allocation or risk rules call for it.
How do I get my updated portfolio?
Strategy monitoring shows the current target holdings, weights and changes since the previous review. Portfolio-change notifications alert you to updates. You can implement those changes yourself, or have your connected AI compare the targets with your brokerage holdings and prepare orders for your approval.
Do I need to run discovery again for every update?
Routine portfolio updates apply the existing strategy rules to new data and do not require another discovery run. Use discovery again to test a new thesis, revise your constraints or evaluate alternative rules.
How should I read a backtest?
A backtest shows how a defined rule set would have behaved over a historical period under stated assumptions. Compare returns alongside drawdowns, volatility, costs and the benchmark. Review data limitations and sensitivity to assumptions. Historical results do not establish future performance or whether a strategy is appropriate for your circumstances.
What is the benchmark line in the demo?
The benchmark is a baseline for assessing return and risk over the same test period. In the demo, the dashed benchmark line is a sample reference series, not a named market index; both curves start at the same value.
Do I have to connect my AI or brokerage?
Both connections are optional in the product. You can monitor the strategy and implement portfolio changes yourself. With an integration, Mirtelo supplies read-only rules, targets and updates to your AI. Your AI connects separately to your brokerage to prepare orders, with execution under your control.
Turn your thesis into a strategy.
Discover portfolio strategies built around your ideas. Tell us the thesis you want to explore first.
You’re on the list.
We’ll write when there’s something real to show. If you left a thesis, it’s already shaping what gets built first.