If you want to build a company, use the time to build relationships with your future customers and credentials for your reputation for being good at whatever yo

If you want to build a company, use the time to build relationships with your future customers and credentials for your reputation for being good at whatever you mean to do. Your employer can't own these. Once you have those relationships, learning their problems and proposing a solution to them, and having them pay for it is also likely outside the scope of your job with your employer. If the problem is valuable enough to them, you can agree to solve it for them for enough money that it will float you for the first 6 months while you develop it. Quit your job, live on your savings, maybe pick up some contracts if you can't raise seed capital, and deliver the product to your customers. Companies aren't made of code, products are, and there are no products without customers, as otherwise it's just art or a science project. Companies are made primarily of relationships with customers, underwritten by products, some of which are made of code. That direction is not reversible. I've developed my own software projects, and for a long time tried to figure out how practising something I could pay someone offshore to do for $30k/year could somehow be worth so much more if I could just do some secret thing right. The advice was always scattershot, be more this, do more that. What it came down to was, it's not the code, it's not the features, it's not the design or architecture - it's the relationships. Who do you have relationships with and what are their problems? If you can't solve them with what you are good at, maybe you need to find new people with new problems. If you can't do that, find someone who knows people who have problems and money, and then partner with them to sell a solution. It sounds so simple, and it is, but if it were easy, everyone would do it.