I created a new consulting business for improving relevance of site search results for say e-commerce, etc. At the time my company did a mix of big data and search consulting.
I took about half my hours to write a book (Relevant Search) along with 20 of my own. I created a tool Quepid (http://quepid.com) to help optimize relevance for Solr and Elasticsearch. Along with colleagues, I started a conference: Haystack (http://haystackconf.com) focused on the domain. I also made an open source plugin for integrating machine learning capabilities into Elasticsearch, and got Wikimedia to partner building it out further.
All that fed the sales funnel for the services I was building out somewhat separate from the companies main business. As the demand picked up, we matured the products we offered in that market.
Eventually the companies focus completely changed to what we were doing and away from the old model.
I will say, while the company would probably try to support anyone doing this, ongoing support depended on results and momentum. I was lucky to have some success, so got to keep building after early proof of market interest.
I took about half my hours to write a book (Relevant Search) along with 20 of my own. I created a tool Quepid (http://quepid.com) to help optimize relevance for Solr and Elasticsearch. Along with colleagues, I started a conference: Haystack (http://haystackconf.com) focused on the domain. I also made an open source plugin for integrating machine learning capabilities into Elasticsearch, and got Wikimedia to partner building it out further.
All that fed the sales funnel for the services I was building out somewhat separate from the companies main business. As the demand picked up, we matured the products we offered in that market.
Eventually the companies focus completely changed to what we were doing and away from the old model.
I will say, while the company would probably try to support anyone doing this, ongoing support depended on results and momentum. I was lucky to have some success, so got to keep building after early proof of market interest.