We continue taking steps forward in ensuring we're positioned to be the #1 automated trading platform across the entire space
Disgustingly large 7 days, another disgustingly large 7 days coming up.
Gunna cover 3 things here:
v2 Metrics Status
Robinhood
Personal Note & Commitment
———-
v2 Metrics - I certainly understated the size of this task. But, the vast majority of the long stuff is done.
Where we're currently at:
- 100+ v2 metrics are currently live and being ingested
- All of which containing health tags
- Health tags update live and inform users on their status to communicate better around their performance in delivering information
- We now ingest data in real-time, on every single trading event. Instead of in longer time buckets (1min+)
- The algorithm simulator now too correlates 1:1, backtesting data in real-time on every single trading event, before it would be at different time buckets or drift in time delay causing inaccurate results. Solved.
- In the process of matching up old metric names to be 1:1 with how they are now on v2. The new ones are named to be much more understandable.
Issue: When using the algorithm simulator on longer time frames with v2, you may see inaccuracies. I am doing my absolute best and trying various different methods to patch up historical data to match 1:1 with the current, but it is proving to be near-impossible to provide the 100% accuracy we're looking for, as it stands, i'd say it's at 50% if you used timeframes from back in August.
As of September 2nd Onwards - it should be over +99% accurate to what you see and what you receive..
I'll keep you posted on success in historical dates, but please take this into account. I will place a notice on the algorithm simulator page.
There were user complaints about most metrics being 'quality related' and little/if any, being time related. I am looking to have these ingesting and live on the platform by the end of today.
- Average Position Size
- Average Position Size (Humans)
- Average Position Size (Bots)
- Human Trader Patience
- Profitable Human Share
- Profit Concentration
- Human Money Velocity
- Market Cap
- Market Cap Change (10m)
- Net Buy Volume
- Price (SOL)
- Total Supply
- Token Decimals
- Launch Liquidity
- Highest Price Seen
- Net Sell Volume
- Market Cap at 15m / 30m / 45m / 60m
- Net Buy Volume at 15m / 30m / 45m / 60m
- Net Sell Volume at 15m / 30m / 45m / 60m
These should allow you to determine well with lots of selection on the type of timing you'd like your algorithm to be signalling. Ingestion would probably begin here on 4th September, take that into account for the algorithm simulator. I'll note that too on the platform so it's clear.
Please over the next few days expect a few the odd hiccup & discrepencacy here and there. I'm aware of all of them and acting on all of them.
———-
Robinhood:
Simply put - the market is going wild there, nobody has even attempted to provide fully automated trading there yet. The opportunity is amazing.
It felt wasted to not continue in rolling the ball there, there is a ton of volume & eyeballs.
I've continued on with Robinhood development whilst tackling v2 work.
Completed:
- Wallet sign-in
- Visual algorithm builder using the live signals
- Backtesting against real historical chain data
- Publish an algorithm to a marketplace
- Browse published algorithms
- Token pages with live price charts and historical candles
- AutoTrader Terminal
- Public signal-health page showing which metrics are trustworthy
- 48 metrics, live & ingesting as of today, 4th September.
What remains:
- Live trading
- Algorithm pages
- Remaining metrics (want to launch at 75+)
I recognize the rush here & impact this would have in launching ASAP.
To speak candidly, there is a shit ton of money & eyeballs on the table here, no promises but I shall try get this out at some point later next week.
———-
Finally to face things head on - I think this v2 release is the first time we've seen eyebrows raised from the community, almost unanimously.
Firstly, to be absolutely clear - this update is a MUST for Occam to scale properly. Simply put, old data was too inconsistent and dirty. No chance of scaling to 1000+ daily concurrent users with that.
I'm not building Occam to stay at 50+ daily users or even 250+.
The addressable market and usecases here are so much larger. If it means a turbulent week in order to lift the covers on what we can achieve, then i'd do it 100x over. Under fire of criticism or not.
I have put the best part of 2 years now into Occam, with 5 years of solely building automated trading solutions in specifically onchain crypto markets. And over 10+ years of building and deliver crypto related software programs around markets, data & trading.
Occam has now seen over 7figs of spend.
For those that have seen how i've worked before, you'll know this - but for those that haven't - i'm nowhere near happy sitting at prices these levels, and i'll work my ass off to ensure we're nowhere near these price levels, as soon as possible.
I love building, but my favourite thing is growth & marketing related stuff. I'm itching to get to that.
Step forward every day, hit me with any and all criticism, feedback, questions whenever you want.
I'll always respond. If I don't, i'm acting on them and building out solutions.
Alright that's enough yapping
Back to work - lots to do as always!