Software developer & investor. CEO at Southport Technology Group. Partner at Southport Ventures. MBA from London Business School, MBA from Columbia Business School.
You'll want to target the side that has bigger pain as your entry point.
In addition, it's also worth considering a narrower subset in the "pain" category. One where you can imagine the onramp process.
An example would be airline reservation system (API). There are ticket buyers and customer service representatives. If you imagine the customer service representatives to have more pain (i.e. a poor experience and lack of features), then you would want to start with that side. A subset of the industry may be smaller regional or budget airlines. They may be incentivized to try something new based on cost concerns and a smaller overall footprint for their products.
You want to build a pain killer, not a vitamin: https://gettacklebox.medium.com/customer-interviews-so-you-ve-decided-to-take-your-startup-seriously-65715f4e9aec
I've always been of the opinion that small business IT services is a very low barrier to entry. Just by virtue of being digital natives (not guaranteed, but likelier than older people), the young people probably have enough of the skills they need to get started.
There are still an uncomfortably high-level of small businesses (and non for profits) that have little to no help with IT services. Basic problems like making sure everyone can connect to the local printer, computers are running antivirus software, and everyone is viewing data with the most up to date browsers. I have been working with an adoption agency in my state, and they lack all the basics: their own domains, any notion of IT security, document sharing, etc.
Given your short timeline and large group, you would be well positioned to partner with small businesses and small non-for-profits in the community. It would make a big impact and produce achievable results. Larger IT projects are hard to get any progress in less than 6 months, even when you have experienced staff. The problems are BIG, and therefore require a lot of implementation and follow-on analysis.