Month: July 2020

Back in 2016, Mobalytics wowed the judges at Disrupt SF with its data-based coach for the exploding competitive gaming world, winning the Startup Battlefield. The company is building on the success of the past few years with a new funding round and a compelling new collaboration with Tobii that uses eye-tracking to provide powerful insights into gamers’ skills.

Mobalytics began with the idea that, by leveraging the in-game data of a competitive e-sport like League of Legends (LoL), they could provide objective feedback to players along the lines of how fast or effective they are in different situations. Quantifying things like survivability or teamplay provides an analogue to similar measures in physical sports.

“On an athlete you have all these measurements, like pulse oximeters, ECGs, the 40-yard dash,” said Amine Issa, co-founder and “Warchief of Science.” Not so much with PC games. Their challenge at that time was to take the LoL API provided by Riot and transform it into actionable feedback, which the company’s success in the years since suggests they managed to do.

But Issa had always wanted to use another, more direct and objective measurement of a gamer’s mental processes: eye tracking. And last year they began an internal project to evaluate doing just that, in partnership with eye-tracking hardware maker Tobii.

“If you know where someone is looking, it’s the closest thing to knowing what they’re thinking,” Issa said. “When you combine that with the larger picture you can put together something to help them along. So we spent six months conducting research, taking players of different levels and roles and studying their eye tracking data to find some metrics we could organize the platform around.”

Not surprisingly, there are characteristics of the highly skilled (and practiced) that set them apart, and the team was able to collect them into a set of characteristics that any player can relate to.

Well, the gif compression isn’t so hot, but you get the idea – the purple square indicates attention. Image Credits: Mobalytics

“We had to think about how to build a product that people want to use. One thing we learned after TechCrunch is that even a simple score from 0-100 doesn’t work for everyone. You need to provide the context for that. So with something like eye tracking, you’re getting 30 data points per second — how do you break that down in a way that players understand it?”

Talking to professional gamers and coaches during the study helped them form the main categories that Mobalytics now tracks with the aid of a Tobii device, like information processing, map awareness, and tunnel vision.

“It’s important to be able to tell a narrative to people. Say you get ganked a lot,” said Issa, referring to the unfortunate occurrence of being picked off by enemy players while alone. “Why are you getting ganked? If your vision score is high but map awareness is low, that’s one thing. Did you know all the information and go in arrogantly, or were you not aware? League is a very complicated game, so players want to know, in this specific fight, what did I do wrong, and what should I have done instead?”

That second question is a tougher one (though perhaps AI MOBA players may have something to say about it), but the metrics are powerful in and of themselves. “Pros are fascinated by this technology,” Issa said. “There’s a lot of ‘I had no idea’ moments. Coaches have said, these are my fastest players but it’s cool to see that as a quantifiable variable.”

A post-game dashboard lets you know your strengths and weaknesses.

Tobii’s head of gaming, Martin Lindgren, echoed this feeling: “Pro teams aren’t interested in being told what to do. They want the data so they can draw their own conclusions.”

Tobii now has a gaming-focused eye-tracker and integrates with a number of AAA games, like Rise of the Tomb Raider, where it can be used in place of fiddly aiming using the analog sticks. As someone who’s bad at specifically that part of games, this is attractive to me, and Lindgren said opportunities like that are only increasing as gaming companies embrace both accessibility and try to stand out in a crowded market.

The companies have worked together to improve the eye-tracking coaching, for instance lowering the number of games a user must play before the system can accurately track their in-game actions; Lindgren said the collaboration with Mobalytics is ongoing — “definitely a long-term partnership” — in fact Tobii’s relationship with the founders predates their startup.

Image Credits: Tobii

The ultimate goal of the Mobalytics is to have a gaming assistant that adapts itself to your playing and preferences, making intelligent suggestions to improve your skills. That’s a ways off, but the company is getting the hang of it. Its first product the LoL assistant, took a year to build, Issa said. A more recent one, for Legends of Runeterra, took three months. Teamfight Tactics took three weeks.

Admittedly it was more difficult to design one for Valorant, which being a first-person shooter is wildly different from the other games — but now that it’s done, a lot of that work could be applied to an assistant for Counter-Strike or Overwatch.

Expansion to other games and genres is the reason for raising an $11M series A, led by Almaz Capital and Cabra VC, with HP Tech Ventures, General Catalyst, GGV Capital, RRE Ventures, Axiomatic and T1 Esports participating.

“It was a very different experience from the post-TechCrunch one, where you’re in the spotlight and everyone’s throwing money your way,” said Issa. “But we’ve built a successful product on LoL, expanded to four games, today we have more than seven million monthly active users… Our plan is to double down on what’s worked for us and create the ultimate gaming companion.”

Source: Tech Crunch

For the first few months it was operating, Shelf Engine, the Seattle-based company that optimizes the process of stocking store shelves for supermarkets and groceries, didn’t have a name.

Co-founders Stefan Kalb and Bede Jordan were on a ski trip outside of Salt Lake City about four years ago when they began discussing what, exactly, could be done about the problem of food waste in the US.

Kalb is a serial entrepreneur whose first business was a food distribution company called Molly’s, which was sold to a company called HomeGrown back in 2019.

A graduate of Western Washington University with a degree in actuarial science, Kalb says he started his food company to make a difference in the world. While Molly’s did, indeed, promote healthy eating, the problem that Kalb and Bede, a former Microsoft engineer, are tackling at Shelf Engine may have even more of an impact.

Food waste isn’t just bad for its inefficiency in the face of a massive problem in the US with food insecurity for citizens, it’s also bad for the environment.

Shelf Engine proposes to tackle the problem by providing demand forecasting for perishable food items. The idea is to wring inefficiencies out of the ordering system. Typically about a third of food gets thrown out of the bakery section and other highly perishable goods stocked on store shelves. Shelf Engine guarantees use for the store and any items that remain unsold the company will pay for.

Image: OstapenkoOlena/iStock

Shelf Engine gets information about how much sales a store typically sees for particular items and can then predict how much demand for a particular product there will be. The company makes money off of the arbitrage between how much it pays for goods from vendors and how much it sells to grocers.

It allows groceries to lower the food waste and have a broader variety of products on shelves for customers.

Shelf Engine initially went to market with a product that it was hoping to sell to groceries, but found more traction by becoming a marketplace and perfecting its models on how much of a particular item needs to go on store shelves.

The next item on the agenda for Bede and Kalb is to get insights into secondary sources like imperfect produce resellers or other grocery stores that work as an outlet.

The business model is already showing results at around 400 stores in the Northwest, according to Kalb and it now has another $12 million in financing to go to market.

The funds came from Garry Tan’s Initialized and GGV (and GGV managing director Hans Tung has a seat on the company’s board). Other investors in the company include Foundation Capital, Bain Capital, 1984 and Correlation Ventures .

Kalb said the money from the round will be used to scale up the engineering team and its sales and acquisition process.

The investment in Shelf Engine is part of a wave of new technology applications coming to the grocery store, as Sunny Dhillon, a partner at Signia Ventures, wrote in a piece for TechCrunch’s Extra Crunch.

“Grocery margins will always be razor thin, and the difference between a profitable and unprofitable grocer is often just cents on the dollar,” Dhillon wrote. “Thus, as the adoption of e-grocery becomes more commonplace, retailers must not only optimize their fulfillment operations (e.g, MFCs), but also the logistics of delivery to a customer’s doorstep to ensure speed and quality (e.g., darkstores).”

Beyond Dhillon’s version of a delivery only grocery network with mobile fulfillment centers and dark stores, there’s a lot of room for chains with existing real estate and bespoke shopping options to increase their margins on perishable goods as well.

Source: Tech Crunch

As AI and machine-learning tools become more pervasive and accessible, product and engineering teams across all types of organizations are developing innovative, AI-powered products and features. AI is particularly well-suited for pattern recognition, prediction and forecasting, and the personalization of user experience, all of which are common in organizations that deal with data.

A precursor to applying AI is data — lots and lots of it! Large data sets are generally required to train an AI model, and any organization that has large data sets will no doubt face challenges that AI can help solve. Alternatively, data collection may be “phase one” of AI product development if data sets don’t yet exist.

Whatever data sets you’re planning to use, it’s highly likely that people were involved in either the capture of that data or will be engaging with your AI feature in some way. Principles for UX design and data visualization should be an early consideration at data capture, and/or in the presentation of data to users.

1. Consider the user experience early

Understanding how users will engage with your AI product at the start of model development can help to put useful guardrails on your AI project and ensure the team is focused on a shared end goal.

If we take the ‘”Recommended for You” section of a movie streaming service, for example, outlining what the user will see in this feature before kicking off data analysis will allow the team to focus only on model outputs that will add value. So if your user research determined the movie title, image, actors and length will be valuable information for the user to see in the recommendation, the engineering team would have important context when deciding which data sets should train the model. Actor and movie length data seem key to ensuring recommendations are accurate.

The user experience can be broken down into three parts:

• Before — What is the user trying to achieve? How does the user arrive at this experience? Where do they go? What should they expect?
• During — What should they see to orient themselves? Is it clear what to do next? How are they guided through errors?
• After — Did the user achieve their goal? Is there a clear “end” to the experience? What are the follow-up steps (if any)?

Knowing what a user should see before, during and after interacting with your model will ensure the engineering team is training the AI model on accurate data from the start, as well as providing an output that is most useful to users.

2. Be transparent about how you’re using data

Will your users know what is happening to the data you’re collecting from them, and why you need it? Would your users need to read pages of your T&Cs to get a hint? Think about adding the rationale into the product itself. A simple “this data will allow us to recommend better content” could remove friction points from the user experience, and add a layer of transparency to the experience.

When users reach out for support from a counselor at The Trevor Project, we make it clear that the information we ask for before connecting them with a counselor will be used to give them better support.

Image Credits: Trevor Project (opens in a new window)

If your model presents outputs to users, go a step further and explain how your model came to its conclusion. Google’s “Why this ad?” option gives you insight into what drives the search results you see. It also lets you disable ad personalization completely, allowing the user to control how their personal information is used. Explaining how your model works or its level of accuracy can increase trust in your user base, and empower users to decide on their own terms whether to engage with the result. Low accuracy levels could also be used as a prompt to collect additional insights from users to improve your model.

3. Collect user insights on how your model performs

Prompting users to give feedback on their experience allows the Product team to make ongoing improvements to the user experience over time. When thinking about feedback collection, consider how the AI engineering team could benefit from ongoing user feedback, too. Sometimes humans can spot obvious errors that AI wouldn’t, and your user base is made up exclusively of humans!

One example of user feedback collection in action is when Google identifies an email as dangerous, but allows the user to use their own logic to flag the email as “Safe.” This ongoing, manual user correction allows the model to continuously learn what dangerous messaging looks like over time.

Image Credits: Google

If your user base also has the contextual knowledge to explain why the AI is incorrect, this context could be crucial to improving the model. If a user notices an anomaly in the results returned by the AI, think of how you could include a way for the user to easily report the anomaly. What question(s) could you ask a user to garner key insights for the engineering team, and to provide useful signals to improve the model? Engineering teams and UX designers can work together during model development to plan for feedback collection early on and set the model up for ongoing iterative improvement.

4. Evaluate accessibility when collecting user data

Accessibility issues result in skewed data collection, and AI that is trained on exclusionary data sets can create AI bias. For instance, facial recognition algorithms that were trained on a data set consisting mostly of white male faces will perform poorly for anyone who is not white or male. For organizations like The Trevor Project that directly support LGBTQ youth, including considerations for sexual orientation and gender identity are extremely important. Looking for inclusive data sets externally is just as important as ensuring the data you bring to the table, or intend to collect, is inclusive.

When collecting user data, consider the platform your users will leverage to interact with your AI, and how you could make it more accessible. If your platform requires payment, does not meet accessibility guidelines or has a particularly cumbersome user experience, you will receive fewer signals from those who cannot afford the subscription, have accessibility needs or are less tech-savvy.

Every product leader and AI engineer has the ability to ensure marginalized and underrepresented groups in society can access the products they’re building. Understanding who you are unconsciously excluding from your data set is the first step in building more inclusive AI products.

5. Consider how you will measure fairness at the start of model development

Fairness goes hand-in-hand with ensuring your training data is inclusive. Measuring fairness in a model requires you to understand how your model may be less fair in certain use cases. For models using people data, looking at how the model performs across different demographics can be a good start. However, if your data set does not include demographic information, this type of fairness analysis could be impossible.

When designing your model, think about how the output could be skewed by your data, or how it could underserve certain people. Ensure the data sets you use to train, and the data you’re collecting from users, are rich enough to measure fairness. Consider how you will monitor fairness as part of regular model maintenance. Set a fairness threshold, and create a plan for how you would adjust or retrain the model if it becomes less fair over time.

As a new or seasoned technology worker developing AI-powered tools, it’s never too early or too late to consider how your tools are perceived by and impact your users. AI technology has the potential to reach millions of users at scale and can be applied in high-stakes use cases. Considering the user experience holistically — including how the AI output will impact people — is not only best-practice but can be an ethical necessity.

Source: Tech Crunch

Meet Diaspora Ventures, a new VC fund based in the U.S. founded by two partners who grew up in France but have been in the U.S. for more than a decade — Ilan Abehassera (pictured right) and Carlos Diaz (pictured left). As the name suggests, Diaspora Ventures wants to invest in the French diaspora, and especially French founders who want to create a startup in the U.S. from the early days of their companies.

The fund’s website lays out this investment thesis in just a few sentences. “We are convinced that France is full of talented and ambitious entrepreneurs and is home of some of the best engineers and product designers in the world. We have realized that most of the time, they do not have the opportunity to expand beyond their borders because they don’t have access to the right funding, talent pool, playbooks or network,” the website reads.

Ilan Abehassera has been an entrepreneur for a while. He first founded Producteev, a task-management product that has been acquired by software giant Jive in 2012.

He later launched Ily, a family-friendly communication device. The device looks an awful lot like an Amazon Echo Show, but the device was announced a couple of years before Amazon took over this market. He’w now working on Willo, a robot that could replace your toothbrush

Carlos Diaz co-founded Carlos Diaz, a European digital agency that was acquired by Atos. More recently, he’s been working on The Refiners, a San Francisco-based seed fund that helps international founders get started in Silicon Valley. In March, Diaz was in the process of raising a second fund for The Refiners but couldn’t close the deal.

“The original idea of The Refiners was to identify ambitious European startups willing to move to the U.S. and help them become global leaders in their category. The global pandemic, the economic uncertainty combined with travel restrictions, and the absurd immigration policies of the Trump administration suddenly made the investment thesis of The Refiners impossible to execute four years after its launch,” Diaz wrote in a blog post.

With Diaspora Ventures, Diaz and Abehassera want to differentiate themselves from French VC funds that already invest in the U.S. According to Diaz, working with a French fund in the U.S. requires a lot of back and forth to close a deal. Diaspora Ventures wants to be able to close deals more quickly.

Diaspora Ventures will focus on early stage rounds with an average check between $100,000 and $200,000. The firm wants to participate in 10 to 20 deals per year.

Interestingly, Diaspora Ventures is taking advantage of AngelList’s Rolling Venture Funds. It means that the two partners have raised $3 million from various limited partners, such as Kima Ventures, Breega, Alexis Bonillo, Christophe Courtin, Salomon Aiach, Frédéric Laluyaux and others. But the fund is always raising, so the list will become longer and the total amount of capital raised will grow over time.

That’s how Diaspora Ventures managed to close their first investment deal just a couple of months after coming up with the idea for the new fund.

Both of them have also been active angel investors over the past few years. They have invested in some well-known names, such as Algolia, Sunrise, Tempow, Yolo, Double, Cowboy, etc.

Image Credits: Diaspora Ventures

Source: Tech Crunch