BY Hannah Adler 3 MINUTE READ is paving the digital way for the events industry. At the touch of a button their platform allows users to arrange a complete conference –  completing the jobs of hundreds in seconds. How does an idea like this originate? How would one implement it? We talk with one of the founders, Deepak Jhavar to discuss why the world needs its first AI-driven conference platform.

Where did the idea for an AI-driven conference originate? What was the gap you noticed in AI and automation?

I ran a B2B conference company from 2000 to 2006 and quit to join a Non-Banking Financial Company (NBFC). It’s a company that does not have a full banking license or is not supervised by a national or international banking regulatory agency. When I quit my job, I took stock of how technology changed our lives when connecting with old friends in the event space. I realised that the

The event industry was using the same processes such as excel, cold calling, paper records etc. from when I started in the industry back in the early 2000s. That was a eureka moment. For the most part, the events industry hadn’t really evolved with today’s tech capabilities. That was the opportunity I believe we’ve identified. We started by creating a detailed process flow chart and established two rules of thumb: first, any task, which is easy for humans and difficult for machines, should be done by humans. Second, any task, which is easy for machines and difficult for humans, should be automated. We’re really focused on two areas for now. The first is the sales cycle which is long, usually 3-6 months to sell one event. It has high potential for automation. The second is the relationship with key decision makers which is largely transactional and predictable.

Where does one start when organising a (almost) fully automated conference? Is data manually imported to start the process?

First, we have a look at the human process as it is today. The usual first stage of the process is “production”. It starts with finding conference opportunities within current trending topics in a particular category. This is done by researching various different websites, including those from competitors resulting in a lot of effort into assessing if such a conference would be profitable. The production team then explores venues, potential speakers and the marketing gears then begin to turn. A brief is given to the marketing team which go on to generate leads for potential sponsors and delegates as well as activating a costly sales team.

Those leads are typically run through a cold outreach email marketing campaign (mostly) or through cold calls which is quickly becoming open to full automation. As potential delegates or sponsors are confirmed, they are sent out a registration form (mostly through email – manually) and they make a payment, either through cheque, bank transfer or online payment gateway. Our approach is more direct and simple. We call it the AlgoEvent way by The AlgoEvent tool creates a predictive conference calendar using data sourcing, analytics and an advanced algorithm to assess trending topics in tech (for now), a calendar of events with locations, a list of potential speakers, media partners, sponsors and delegates.

This data is then fed to an automated email and social outreach campaign for attendee and sales acquisition. The only human intervention in the process remains in the area of sales calls. Suppliers are also organised by the AlgoEvent tool ensuring smooth, automated logistic management and payment. Manually, the cycle time is about 5-6 months and also very inconsistent across the globe. This has huge ramification to the end result which we can now do at a click of a button.

Adding to the above question, from where did the AI’s algorithm gain its knowledge?
Were similar conferences studied for their overall performances?

Our proprietary algorithm is backed with data of the last 6 years of similar conference companies. Taking into account the number of tech conferences happening every single day across the globe, there are more than 300. The knowledge the model gains out of this is unbelievably huge.

The Blockchain Investment Roadshow in Cape Town will be an event of about 500-600 people. What is the limit to Pronoia’s abilities in organising such an event?

Our vision at is to do the same task more efficient with minimum human intervention. To be honest, there is no limit to the number of audience we can invite. In a couple of years, after we have trained our model enough, we can do close to 250-300 conferences in a year with an average of 400-500 delegates per conference with minimum human intervention. AI technologies are now present in almost every aspect of our lives.

What are your opinions on recent fear-mongering of AI in general?

AI is here to stay. Let’s take a case of Uber: did it take away jobs or did it create opportunity? In my view, the profile of jobs might change due to the of AI.