There are always ways to innovate within any extent system, even systems that resist innovation. For example, in USA, the current President is innovating by declaring his candidacy for the 2020 elections very, very, very early. This suppresses the degree to which non-profits can criticize him, and bestows fundraising advantages. For better or worse, he operates entirely in the space between campaigning and governing, where one might exercise "biz ideas," and exploiting/exploring the boundaries of the election system is a bizidea.
But, the people can innovate as well. What if some polities in your country instituted at-large Range Voting? (rangevoting.org) What if candidates could strategically pick a constituency? Say, the at-large candidate whose issue is advancing the protections against rampant automation? Or the at-large candidate whose mobile app was wildly popular among an unexpected demographic group, like Walmart workers?
Here's an idea for iterating citizen policy making: You can start a simple web site that generates unambiguous and approachable names for legislative acts. At first, the act is simply vapor; your site asks individuals to contribute what they think such a bill contains. Eventually a mass of people might tune it to be as publicly attractive as possible. I'll propose something random: The Alligator Farm Protection Act of 2017. It starts out completely blank and your site encourages people (who're interested enough to click) to contribute some editing to the text. Maybe the text won't become legalese, which puts off many. Maybe you'll have an NLP feature where different contributors' changes are merged intelligently to preserve sentiment. Multiple contributors make this an iterative process. Once you've iterated to a point of average acceptance, maybe the Flux party web site will sponsor your bill for a parliamentary vote.