Congrats on the launch!
Launch HN: Jasmine (YC S22) – Automating REC compliance and payouts for solar
11–20 of 22 posts
Re: Launch HN: Jasmine (YC S22) – Automating REC compliance and payouts for solar
#12Have you reached out to the folks at https://gosolarapp.org/ to see if there are any potential partnership or integration opportunities? Congrats on the launch!
Re: Launch HN: Jasmine (YC S22) – Automating REC compliance and payouts for solar
#13Re: Launch HN: Jasmine (YC S22) – Automating REC compliance and payouts for solar
#14Re: Launch HN: Jasmine (YC S22) – Automating REC compliance and payouts for solar
#15Re: Launch HN: Jasmine (YC S22) – Automating REC compliance and payouts for solar
#16Re: Launch HN: Jasmine (YC S22) – Automating REC compliance and payouts for solar
#17This is really great! I have solar on my roof at home but am not sure who owns the RECs - me, the installer, the utility company, rando Jon Doe? Does your platform provide a method to confirm that as well?
Sometimes that's explicitly in the documents and other times it's based on the jurisdiction of the solar installation.
Re: Launch HN: Jasmine (YC S22) – Automating REC compliance and payouts for solar
#18Congrats on the launch!
Re: Launch HN: Jasmine (YC S22) – Automating REC compliance and payouts for solar
#19This is amazing and so necessary to drive down the cost of solar and speed up its adoption. Godspeed!
Re: Launch HN: Jasmine (YC S22) – Automating REC compliance and payouts for solar
#20> This domain is a good fit for automation and LLMs—not to generate text, but to (1) structure unstructured documents, (2) interact with legacy government websites where there’s no API, and (3) deal with repetitive bureaucratic language.
This isn't a criticism of what you're doing, but a more general gripe/musing about the wider software and AI ecosystem. I've seen this in my own work too. I feel very unhappy that we are using complex, nondeterministic, power-hungry "intelligent" machines to solve the problem of... unstructured data. Instead of... structuring the data.
I know you can't solve that problem. But nevertheless, wouldn't it be better for society as a whole if "we" agreed to make data accessible in machine-readable ways that don't require human-like agents to piece together the mess?
This is a writ-large version of the joke about writing an email in bullet points, inflating it to paragraphs using an LLM, then the receiver summarising the paragraphs back to bullet points using an LLM.