Molly wants to use your online presence to create an automated knowledge base

Isn’t it frustrating when you ask a friend a question — like what’s your favorite restaurant in New York or what trips have you been on this year — knowing that these specific answers are certainly already accessible on social media?

The problem is no one wants to spend an hour combing through their friends’ social media pages (or worse, monitoring them 24/7), so we just end up asking them directly.

Molly wants to fix this. The startup wants to make information more easily accessible, primarily by cross-pollinating information posted on your various online profiles and making it available in one central location. The startup was founded by Chris Messina, Esther Crawford and Ethan Sutin over the summer, and is now part of Y Combinator’s Winter ’18 batch.

Eventually this information can be made available via an Alexa skill or a chatbot, so you could theoretically say “I want to have dinner with Kylie tonight, pick a restaurant we both haven’t been to, but one our individual preferences suggest we’ll both like.” If an automated database can remove the legwork of answering these basic questions (that already have an answer if you just know where to look), more time can be spent on actually interacting and spending time with one another.

Of course, because of natural language processing constraints and machine learning model training time, this full vision is at least a couple of years out, according to Messina. But they have to start somewhere, so the first iteration of Molly is an AMA (ask me anything) feature where audiences can find answers about a certain person, with those answers being aggregated from a wide variety of sources, like Medium, Twitter and Instagram.

Right now they’re only launching with featured profiles (but anyone can ask questions) who have done a Product Hunt AMA in the past, since this is an easy database of questions and answers they can scrape to pre-populate a lot of information. Molly also will occasionally send users quizzes to take, with the answers being recorded in their knowledge bank for future audiences to access.

And lastly there’s a feature for questions that Molly can’t answer to automatically be forwarded along to users, so high-profile people can have a centralized place to take all questions from fans (and not have to answer duplicates, because Molly will automatically find the answer if it has already been asked).

Eventually the plan is to open up profiles so anyone can have a database of answers and preferences for friends to access. And these don’t all have to be totally public — Messina envisions a potential one-time permission-based system where you could grant a friend access to Molly just for specific purposes and a set period of time, like finding a restaurant for tonight.

Right now Molly’s founders have said it’s too early to think about monetization, and they’re focused on finding product-market fit. But Messina hinted that (way) future versions could use the knowledge base they’ve built up to recommend restaurants or bars they’ll know you like — which could be a future source of revenue.

The startup has raised $1.5 million from BBG, Betaworks, CrunchFund and Halogen Ventures.