As I mentioned in my previous post, 2020 was full of moments were I was ashamed of not giving enough time as a CEO or as a DAD...
To better understand how I handled this, let's start by looking at where explor.ai's business idea came from.
The business idea - AI As A Service
I worked on explor.ai's business plan for the most part of 2019. As a Chief Data Officer at Biron Health Group, and had multiple pain points I wanted to fix :
- Access to full time or part-time AI expert talents was really hard
- Business Intelligence experts need sporadic help on their AI / ML projects
- AI partners/companies willing to step in for small (less than 40h) projects were, inexistant
- There is a gap between AI Experts and Software Developers, placing ML/AI solutions deployments somewhere in the middle of their expertises
Operationalizing Data Science
Additionally, I started getting more familiar with solutions and methodologies that can drastically reduce the time spent on a data science project. I call this "Operationalizing Data Science". Breaking down the different steps of data science in sub-expertises, using the proper (modern!) tools, using already proven approaches... Put together as a whole, the system proved capable of reducing the cost of projects, opening up the possibility to tackle some projects in as little as 5 hours.
Adding in the recent development of AutoML tools, I felt explor.ai had the recipe for a service offer that could drastically lower the cost of AI projects in general.
Using AI to build AI
Our mission statement (which admittedly is still a work in progress in 2021!), is to
Apply artificial intelligence on data science processes to reduce the initial costs of new projects, optimize the time of data scientists / AI experts and make it easier to turn a positive return on investment on short projects
I have a fair amount of experience on operations and processes in general, and have put a lot of efforts on minimizing the costs normally associated with launching and deploying machine learning systems for clients. That is, we reduced friction between the client's team and explor.ai's team in a way that reduces project overall costs.
AutoML tools are a significant part of this system and, incidentally, explor.ai developed quite a bit of expertise in that field!
