My initial plan was to use a Raspberry Pi to connect the sensor board wirelessly to the internet. The Pi currently has a 433 reciever to get the sensor values from the ATTiny and is connected via LAN to the internet.
This has several issues. A) the Pi is rather expensive and overpowered to just connect something to the internet and host a simple web server. B) Cheap 433Mhz transmitter/reciever suck donkey balls regarding range, especially indoors.
I really didn't want to spent much money on the project when I started out so I opted for the cheapest solutions possible. I changed my mind a bit as the project was interesting to do for me so far (I'm still a huge noob with electronics) and I want to make a better hardware but with a low price still in mind.

So my plan now is to atleast ditch the 433Mhz transmitter and opt for LAN or a NRF24l01+ 2.4GHz transmitter, which have a better range when properly configured.
My preferred solution would be to also ditch the Pi and go for a "VoCore" which I recently saw on Hackaday. It's a 20$ Wifi enabled MIPS board that can run OpenWRT. The plan would be to  directly put this cute little guy on the board and connect it to the ATTiny via serial/spi/i2c.

I have already written a huge part of the server in Node.js aswell as the client front end. The client communication and user management is already done, what is still missing is the sensor value analyzing to determine if a machine is running. Which is more difficult than expected as the sensor board is not exactly accurate.

I will probably have to redo some stuff in order to get a more accurate reading. It's absolutely possible to tell that a machine is running but with a delay of ~15-20 minutes which is not acceptable. It also has possible "false alarms", sometimes the ADC reading spikes without anything happening or sometimes it misses an important spike when starting a machine which leads to the said delay.

I pretty much copied the CT sensor part from the great Open Energy Monitor project (http://openenergymonitor.org) and didn't entirely understood their circuit to convert the amps from the ct sensor into a voltage. Now that I have a better understanding I hope I can improve the accuracy of the readings.