Wellwishers' Newsletter, December 2016


Since our October newsletter, version 1 of the Legalese system completed three more deals: two for users … and one for ourselves! That’s right, Legalese just used Legalese to engross over S$600,000 in hard commitments from angels and VCs. Eating your own dog food: delicious.

Our software generated nearly dozen documents, signed over four stages. Our v1 product produced everything from directors’ resolutions and members’ resolutions to term sheets and investment agreements. Negotiated changes were parameterized in Google Spreadsheets and tracked in Git; Legalese turned XML into PDFs and them directly into Adobe EchoSign for signature.

The best part? Legal fees: $0.

If we had gone to a white-shoe law firm, what would we have had to pay? Assuming they’d even be willing to take us as a client, we might be out of pocket anywhere between $2,000 and $20,000. No wonder entrepreneurs are desperate for a solution.

But will they pay? How will they pay? Burning questions – which the $600,000 will help to answer over the months to come. Presenting at Harvard Law School, Meng suggested that Legalese might commoditize contracts all the way down to utility pricing – and nobody blinked. Imagine a future where clients pay a base fee of $1 per signature, irrespective of the subject matter of the contract. Or a flat retainer of $1000 a month to manage all the contracts in an enterprise. If Hubspot and Salesforce can charge that kind of money on CRM, surely we can do the same for contracts, a mission-critical system of record that still runs on human labour.

In our last newsletter, we forecast that we would:

  • continue v2 specification, architecture, and development
  • ship a term sheet and definitive documentation to investors for review, and
  • prepare to present a draft of our very early work on L4 at Geekcamp Singapore (Oct 24) and at the Hackers Conference (Nov 5).

We have done most of these things.

Our outsourced dev shop is now working daily on v2 use case elicitation, storyboard wireframing, and system architecture.

Due to conference craziness we weren’t able to get on the docket for Geekcamp Singapore. But we did present Legalese at Hackers and at Harvard.

From Hackers, we are recruiting an experienced full-stack technologist to join the project. I am excited about this, because the last time I presented a major project at Hackers, I found a (different) experienced full-stack technologist who became my partner on the SPF project. Today, SPF (RFC4408, etc) enjoys global adoption as an infrastructure technology standard based on a domain-specific language of my design. Following the same playbook, I hope that Legalese will become another globally adopted technology standard based on a domain-specific language of my design.

On the research front:

Legalese’s programming language, codenamed L4, will be to legal contracts what SQL is to databases, what Verilog is to chip design. Needless to say, designing a new programming language is one of the biggest adventures in computer science. L4 will take years to develop, and we’ll need help from academia to do some of the heavy lifting. To this end, we have started exploring grants under a couple of Singapore-based research and commercialization schemes: one for $250,000, and another for $5M. These grants will fund use-inspired academic research. But we are new to the academic grantwriting world, so we don’t expect to get it right the first time.

That said, the last few months have been fruitful: two members of the Legalese team have been in Cambridge, MA working on DSL design, and they have produced formalized a simple convertible note in L4, extractable to the CSL language. More on this in our next newsletter.

On the product front:

Over the last few months, the v1 system produced 139 signatures over 69 documents for end-users – not counting the fundraising for Legalese itself. Supporting end-users requires a lot of work by our staff. In v2, we’re moving that work into software. We expect to have a hands-off v2 system in a few months, as a user-friendly skin running on top of v1. After v2 goes up, monetization experiments will follow. That will tell us how product/market fit is looking. If the fit is good, we will be in a good position to go for a seed round.

On the company front:

At JFDI, this quote hung on the wall: Companies build products. Entrepreneurs build companies. According to Steve Blank, the time for company building will come, after product-market fit. But, keeping in mind the Bene Gesserit admonition “A beginning is the time for taking the most delicate care that the balances are correct”, we’re also laying the foundations for a company that can grow into a billion-dollar category winner. How? By baking in opensource, online-first DNA from the ground up, with management distributed throughout the organization and pushed to the edges. These foundations are already beginning to pay off, as new contributors turn up weekly on our Slack chat and mailing lists.


Between now and the next newsletter, we will:

  • build the first iteration of the V2 product
  • draft a few more legal agreements in L4, thus converging on a language design
  • sort out Legalese’s ESOP structure in a way that can generalize to other startups hire our first few staff members under our hybrid cash/equity plan

Happy Holidays!