This is not an answer but a comment which is way too long for the comments section.
What you can trade will be very much restricted by where you live. In Europe, you have Priips, which changes the game a lot - even US-ETFs are generally not easily available on most platforms (within Europe).
How much money do we actually talk about?. No need to answer this here directly, but that is the first thing that should matter for you (more on that at the very end). If it is sizeable, I think your best bet is to get in touch with some institution that focuses on (U)HNWI. Switzerland, particularly Geneva, seems to be a wonderful place for this. They will find you ways to do (almost) everything you want. It will cost you, but depending on your wealth, the doors they can open, will offset their fees (quickly). For example, Leonteq offers something very flexible. It is not for retail either, but ultimately, becoming a qualified or professional investor is not an impossible hurdle. Or having someone do this on your behalf also works.
Out of interest, what exactly is your triple barrier KI option? The only time where triple barriers make sense (to my knowledge) is with baskets. With one underlying, you could probably get some awkward sequential series but I have never seen anyone ask about something like that. If it is like a Triple Digital (payout if each underlying is above/below a certain barrier at maturity), that is illiquid, and difficult to price. Add your touch (which probably means continuous observation) and you face a real modelling challenge. Why does it matter? Well, your counterparty (if you find one) will cook up a price. If you are retail, and cannot model it, your counterparty will not be kind to you. Also, if you are wrong and the market moves entirely against you, getting rid of a product may actually turn out to be quite difficult (e.g. who is going to buy back your dual currency pivot tarf that turns out to burn money like the Great Fire of London gutted the city).
Assuming you have the knowledge and models to price these things properly, where do you get the data from? You cannot simply use some vendor like Bloomberg, Reuters etc for various reasons:
- Firstly, you sign a so called desktop usage license with the vendor, so you cannot use it anywhere else.
- Secondly, if your firm has a license, they will not be happy for you
to use that data outside of your firm, and doing so most likely
breaches the general agreement between your firm and the vendor
- Thirdly, neither Bloomberg nor Reuters will provide reliable data that you would need for pricing something like a triple digital. Assume it is FX, the SLV model that Bloomberg offers is good, but it calibrates to barrier option prices (check out their white paper if you have access to Bloomberg). These prices are not available for clients. However, this model does not work for baskets anyways - so you only get LV (or if FX LVLC - local vol local correlation which is somewhat better).
- You also cannot (per desktop agreement) use the vendor models directly to price this for personal trading.
Now, assume all this is not an issue, the more you customize, the more the counterparty will charge you (one way or another, through fees, prices etc). For example, due to practical difficulties in replicating the actual log payout across strikes, basic vanilla variance swaps trade at a basis for institutional investors. That basis will unlikely be the same for retail (and that is nothing customized, just plain vanilla VS).
Sometimes, it also helps to think outside the box. If your triple barrier for example is just a down and in basket put (WO, weighted, Rainbow, whatever) - that is a really just a reverse convertible where plenty of products (similar to the products from the RBC example in ne of the comments) are available. While you may not be able to say I want a rainbow basket on these 3 stocks, with this tenor and the down in barrier of x percent, you should at least find something reasonable close.
Last but not least, did you actually test these strategies that you have in mind somewhere? Not losing money (yet alone beating a conventional buy hold of an index) will be very very hard to say the least. The sell side does not trade stuff like this to make money with positions, they make money by offering this service and trying to hedge away directional risk. If you are sure you can do this for a profit, I would say it is probably best to demonstrate this to a (large) hedge fund, and you will soon be rewarded with bonus payments that will almost certainly dwarf your personal trading profits (unless your investable wealth is actually really high, in which case I would recommend asking someone in Geneva to find you a way to trade this).