3 ms·

"Can anyone from Invenia explain how they make money or what they're trying to do?" Whenever you have to ask that question about a tech company, you can genera

"Can anyone from Invenia explain how they make money or what they're trying to do?" Whenever you have to ask that question about a tech company, you can generally assume that the answers are "venture capital" and "attract more venture capital", respectively.

rovolo 7y ago
That's only if they are offering some sort of product. Based on their website blurb, they're probably getting paid as consultants by electric companies: > We interact directly with the grids, helping to plan for generation, flow and use of electricity in advance of real time operations. We help the system operators to optimize the power grid to ensure reliability, efficiency, transparency, while reducing harmful emissions.
- 7thaccount 7y ago
  Each system operator that I know of (I'm not omniscient) has a planning department (1-30 years), as well as groups which handle near term planning (1 day to 6 months out). It wouldn't be unheard of to work with outside consultants on special projects, but that would normally entail working with someone that has very specialized experience in power systems operations (ex: expert in voltage stability). I'm not sure if that is the case here, but maybe one of their employees will stop by and enlighten me. Again, wording is really tough in online forums. I am genuinely curious about the role they serve. Although AI is hardly new in power systems (lots of lisp and prolog applications were created in the early 90s), those were mainly expert systems as opposed to the machine learning of today. Like I said, there are definitely some applications, but usually nothing that some Python or R can't handle. I'm not sure if there are many products that would sell in this space. Maybe their employees have picked up enough domain knowledge coupled with good development skills to help with machine learning projects like you have guessed? It is certainly possible.
  - nyir 7y ago
    If what I've seen is representative, there could be a big opportunity in overhauling the short/medium term forecasting piece in a lot of those companies and adding in a versatile timeseries and analysis platform, which would then allow you to do more accurate forecastings more often (and given the number of readings you're gonna get that might need some thought). And since you're already developing something new you can probably get a nicer architecture out of the whole thing anyway, which might allow you to use all that data in more ways than just for that single purpose. Getting rid of some lock-in from certain database vendors in the long-term might also be nice. Just speculation, I kinda agree in that I doubt yet another optimisation package is gonna be so useful.

rovolo 7y ago

That's only if they are offering some sort of product. Based on their website blurb, they're probably getting paid as consultants by electric companies: > We interact directly with the grids, helping to plan for generation, flow and use of electricity in advance of real time operations. We help the system operators to optimize the power grid to ensure reliability, efficiency, transparency, while reducing harmful emissions.

7thaccount 7y ago
Each system operator that I know of (I'm not omniscient) has a planning department (1-30 years), as well as groups which handle near term planning (1 day to 6 months out). It wouldn't be unheard of to work with outside consultants on special projects, but that would normally entail working with someone that has very specialized experience in power systems operations (ex: expert in voltage stability). I'm not sure if that is the case here, but maybe one of their employees will stop by and enlighten me. Again, wording is really tough in online forums. I am genuinely curious about the role they serve. Although AI is hardly new in power systems (lots of lisp and prolog applications were created in the early 90s), those were mainly expert systems as opposed to the machine learning of today. Like I said, there are definitely some applications, but usually nothing that some Python or R can't handle. I'm not sure if there are many products that would sell in this space. Maybe their employees have picked up enough domain knowledge coupled with good development skills to help with machine learning projects like you have guessed? It is certainly possible.
- nyir 7y ago
  If what I've seen is representative, there could be a big opportunity in overhauling the short/medium term forecasting piece in a lot of those companies and adding in a versatile timeseries and analysis platform, which would then allow you to do more accurate forecastings more often (and given the number of readings you're gonna get that might need some thought). And since you're already developing something new you can probably get a nicer architecture out of the whole thing anyway, which might allow you to use all that data in more ways than just for that single purpose. Getting rid of some lock-in from certain database vendors in the long-term might also be nice. Just speculation, I kinda agree in that I doubt yet another optimisation package is gonna be so useful.