Ask HN: Help me raise a small seed round

7 points by pinouchon ↗ HN
We are now a team of 3 co-founders working full-time on Tufa (http://tufa.ai/). The long term goal is to do AGI research (french version of DeepMind). The short term goal is to provide funding for long term goal by developing an algorithmic trader of cryptocurencies, using deep learning techniques. This has the benefit of developing ML/Deep Learning expertise that will be useful later for research.

So far, we have a basic prototype that makes a monthly 15% of return on capital by trading BTC/EUR on coinbase, as evaluated on backtesting scenarios. We are in the process of deploying it live to validate that our hypotheses hold true.

My brother and I have applied to the Entrepreneur First accelerator and got accepted (https://www.joinef.com/). But we have not signed yet. Their basic deal is a 90k investment for a 10% equity in the form of a convertible note. The issue with their program is that they have a lot of conditions/clauses (first refusal, future investments, dilution, non-compete, intellectual property, board seats) that we are not too happy about. In addition, the contract is phrased in very vague terms.

So I'm evaluating other ways I could raise funds to see if we can raise on better terms. Running on our own savings, we have 2 years of runway that we can extend to 4 years if needed by relocating. The way we plan to use external investment (VC or angel) is as capital for the trader, because there is no need for salaries or offices. This way, we only need a small amount to get started.

We are looking for a deal like this: - $70k worth of BTC for 5% of equity in Tufa, or - $140k worth of BTC for 10% of equity in Tufa, with whatever means is easier (convertible debt, safe, equity...)

Do you know an angel investor that would be interested in this? Thanks

5 comments

[ 2.9 ms ] story [ 18.4 ms ] thread
"... as evaluated on backtesting scenarios. We are in the process of deploying it live to validate that our hypotheses hold true".

It's easy to create an algo (AI or not) that performs well on historical data.

Until "hypotheses hold true" is done and proven - it's better to ask investment based on "we have a such great team and can prove it" vs. "we have such a great idea and hope to eventually prove it".

Nothing is impossible - i just don't think this idea at that stage is an easy sell.

Thanks for the feedback. I think the next step for me is to actually try it out live and see how it performs. If the results are positive, I will have an easier time selling my idea
4 letter .ai domain name. Nice!
"$140k worth of BTC for 10% of equity" -- what makes you think the company is worth $1.4M?

Trading strategies are a dime a dozen, that kind of backtesting does not mean much because there is very limited data, there were historic runs, and there are lots of whales trying to squeeze alpha out of BTC charts.

Then mentioning AGI in the same breath does not give confidence.

That was the minus side.

On the plus side, you guys are inspired and working towards your vision which is awesome!