[Android, iOS] Free - Dana: AI Stock Market Research Tool (20-Agent AI Council, 230 US Stocks) @ Dana.technology

AssociatedLBardz @ dana.technology on 09/09/2026 - 20:38

Moved to Forum: Original Link

Disclosure: I built this. I'm the developer, posting my own app.

DANA is a stock-market research tool. Every week it runs 20 separate AI agents over 230 US-listed stocks — each looks at one angle (momentum, volatility, options pricing, SEC filings, insider and congressional disclosures, ETF flows, analyst revisions) and votes BUY / SELL / HOLD. You see every individual vote and the reasoning, not just a final number.

It is general information and education only. Not financial advice, not a recommendation, and it does not consider anyone's personal circumstances.

What's free — Everything. No paywall, no trial timer, no card. I built paid tiers earlier and switched them off. I'm not promising it stays free forever; if that changes I'll say so.

What it does — Weekly signals with the full vote breakdown · 10 years of frozen, auditable past signals · paper trading · backtesting with transaction costs · a world map layer for macro context · AI blog and daily podcast.

What it does NOT do — No broker connection, cannot place trades, never touches your money, doesn't tell you what to buy. No ads, no data selling.

Caveats — Past performance isn't a guide to future results; historical figures are hypothetical. The model has had bad stretches and they're all visible in the app. US stocks only, no ASX. iOS 26+. Capital at risk — consider a licensed adviser.

Android App:

https://play.google.com/store/apps/details?id=com.dana.tradi…

iOS App:
https://apps.apple.com/au/app/dana-ai-market-research/id6802…

Free Software/Websites Android App iOS App

Related Stores

dana.technology
dana.technology

Comments

Best
Search through all the comments in this post.
  • 191919 on 09/09/2026 - 20:50

    ozslopbargains.com.au

    • 1
  • Wort on 09/09/2026 - 23:12

    Holy glitch, just about every text box is overflowing on the website. Unreadable

    So many contradictory claims, is there 20 agents? 21? 18?

    T and c reference none exist features and charges. So much slop

    • 1
  • brendanm on 09/09/2026 - 23:31

    That website gave me autism.

    • 1
  • life is suffering on 09/09/2026 - 20:51

    Donna ?

    • 0
  • Robin2012 on 09/09/2026 - 20:56

    US market only?

    • 0
    • AssociatedBardz @ dana.technology on 09/09/2026 - 21:12

      For now - I am thinking of doing ASX now but depends on how many users I get. If there is a good response from people ASX can be built as well.

      • -1
  • monky on 09/09/2026 - 20:56

    How do you have 500+ weeks of voting data when agentic workflows are only a couple years old?

    • 0
    • AssociatedBardz @ dana.technology on 09/09/2026 - 21:11

      very easy - the agents were not seeing the future (i.e back testing).

      Simply - there is a vector data base all events on a particular date. Agents were voting based on what they were seeing.

      For example we know what has happened from 01-01-2016 through to 10-12-2021.

      On 10-12-2021, agents looked at everything from 01-01-2016 through to 10-12-2021 and voted on that date and so on without seeing the future.

      • 0
      • tenpercent on 09/09/2026 - 21:38

        Where's the results of the backtesting? Sell it to me.
        So far your pitch is "I have this thing that does X" but you don't show me that it does X well.

        • 0
        • life is suffering on 09/09/2026 - 21:46

          That's proprietary ?

          • 0
          • AssociatedBardz @ dana.technology on 09/09/2026 - 21:56

            @life is suffering: Not really - backtesting is important for the agents to learn

            • 0
        • AssociatedBardz @ dana.technology on 09/09/2026 - 21:55

          all yours:

          https://dana.technology/calculator

          I will stress understanding the past does not necessarily define the future. But in the past 10 years this is what the performance was.

          • -1
          • tenpercent on 09/09/2026 - 22:33

            @Bardz: Thanks. Is there a way to look at the backtesting for specific years? i.e. excluding 2026 calendar year which seems to be where most of the over performance occurs.

            • 0
            • AssociatedBardz @ dana.technology on 09/09/2026 - 23:26

              @tenpercent: Yes sure have a look at backtest page it has the results since 2016

              • -2
              • tenpercent on 10/09/2026 - 11:34

                @Bardz: Was this trained going back further than 2016?
                And can I select a custom date range rather than 'last 1 year', 'last 2 years, … , 'last 10 years'?
                For example I'm much more interested in 2016 - 2025 than 2016 - 2026. Like I said, the graphs seem to show that most of the over-performance vs SPY is just in the last 8 months.

                • 0
          • tenpercent on 10/09/2026 - 11:36

            @Bardz:

            I will stress understanding the past does not necessarily define the future.

            Of course. But if your model / approach can't even outperform the broader market on back testing, then I would be highly skeptical it could going forward.

            • 0
  • Wiadro on 09/09/2026 - 21:25

    this is ozbargain, not ozspam/ozslop

    • 0
Login or Join to leave a comment
Login Join