Julie Desk
side capital

Julie Desk, the email-based virtual assistant that helps you schedule meetings, picks up €600K funding

Next Story

Google+ now gives verified profiles, early access to new features to some Collections and Communities users

Julie Desk, the French startup that’s developed a “virtual assistant” to automate the email back-and-forth typically associated with scheduling meetings and appointments, has picked up a further €600,000 in funding.

The investment comes from SIDE Capital, the VC firm that counts a number of prominent French entrepreneurs as board members, including Frederic Mazzella (Blablacar), Thierry Petit (Showroom privé) and Olivier Mathiot (PriceMinister). SIDE managing partner Renaud Guillerm, who co-founded Vide Dressing, joins the board of Julie Desk.

Launched in early 2015 by the same company behind WePopp, the now shuttered app that attempted to make it easier for friends to plan events together, Julie Desk is an “AI-driven” service for scheduling meetings. Similar to U.S. competitor, it shuns the need for a separate app and instead uses the humble CC: field in your email client as its interface.

Simply CC “Julie” at the start of or during an email thread with the person you want to schedule a meeting with, and, in theory, the virtual assistant will take care of the rest. This includes offering up potential meeting times — pulled in from the various online calendars supported — confirming the agreed time and place, adding it to your schedule and sending out the appropriate calendar invites.

Furthermore, all of this takes place via email, using human-friendly language, as if Julie is a real person. And, actually, sometimes Julie is a real person. That’s because, in some ways, the service is still employing a hybrid model, with its AI being underpinned by human supervision.

“Any email produced by the AI is quality checked by a human before sending, so that we make sure that there are no mistakes and the best quality comes out of Julie,” co-founder and CEO Julien Hobeika told me in an email.

“If something comes out wrong from the AI or simply if the AI can’t produce an answer to a request, the human operator can take over and answer manually. Our clients are business oriented people, they don’t want to waste time and money correcting AI that picked up a word wrong, we do that for them.”

Perhaps human overlords are the future of AI after all. Now that doesn’t sound quite so scary — or does it?