Microsoft’s Cognitive Services Demonstrate Detection of Emotions in Human Face


NEW YORK—Last August 25, Nick Landry, senior technical evangelist at Microsoft, held at demonstration of Microsoft’s Cognitive Services and its 22 APIs (previously called Project Oxford) at Microsoft Reactor at Grand Central Tech.

Imagine an API that could detect the emotion in a person’s face in an image, not to mention tell someone’s age, which became hugely popular when people used the technology from the How Old Do You Look app?, which it is improving; not good news for those who hide their real age.

Now Microsoft is allowing developers to customize the new Cognitive Services. This was highlighted in previous events of Microsoft this year.

The rebranding to Cognitive Services also means that it brings together Bing, Oxford and Translator APIs.

The new Cognitive Services APIs include emotion (comparing facial expressions); entity linking (a textual analysis function); face (facial recognition); linguistic analysis, speaker recognition, speech (speech to text); video (vision analysis); WebLM  (an SDK for the Web Language Model).

Seeing AI shows how these new capabilities can help people who are visually impaired or blind understand who and what is around them. See how it works here

Landry demonstrated how developers can use all these services at the meetup.

Cognitive Services is reportedly a nod to IBM’s Watson, which has been marketed as a “cognitive computing” product, one that’s based on the way the human brain works.

‘Trust is essential in sharing economy’– Sundararajan


By Dennis Clemente

NEW YORK—Trust is essential in the sharing economy.

Last August 17 at Rising Minds, Arun Sundararajan, sharing economy expert and New York University Stern School of Business professor talked about his new book, “The Sharing Economy: The End of Employment and the Rise of Crowd-Based Capitalism” as well as that salient ingredient in how things has worked so far this economy: trust.

If you’re not in this space, it’s really about how your car (for Uber) or your apartment (for Airbnb) and earn from it. For those companies, they make use of the crowd to work for them and if you’re part of the crowd offering any of those services, you get to earn from being part of it. If you have an apartment in Manhattan, for example, you can have it rented out to anyone as long as your landlord approves.

But this only works if you can trust someone to live in your house or apartment or like in Europe, share a ride with strangers. And for some people, they’re willing to sacrifice that to make use of their underutilized assets.

Sundararajan pins down the sharing economy as an economic system with the following five characteristics:

  1. Largely market-based: Enabling the exchange of goods and the emergence of new services, resulting in potentially higher levels of economic activity.
  2. High-impact capital: Opens new opportunities for assets and skills to time and money, to be used at levels closer to their full capacity.
  3. Crowd-based networks rather than centralized institutions or hierarchies: The supply of capital and labor comes from decentralized crowds of individuals rather than corporate or state aggregates; future exchange may be mediated by distributed crowd-based marketplaces rather than by centralized third parties.
  4. Blurring lines between the personal and the professional: The supply of labor and services often commercializes and scales peer-to-peer activities like giving someone a ride or lending someone money, activities which used to be considered “personal.”
  5. Blurring lines between fully employed and casual labor, between independent and dependent employment, and between work and leisure: Many traditionally full-time jobs are supplanted by contract work that features a continuum of levels of time commitment, granularity, economic dependence and entrepreneurship.

The sharing economy has created many thriving businesses for one reason other than trust: Weak regulation, but Sundararajan thinks it should adhere to some regulatory measures, a reinvention of regulation, as most of regulatory interventions have been governed at the local level. Most of these businesses are national, if not international.

If Sundararajan had his way, he would call the sharing economy “crowd-based capitalism,” because the former still leaves a fair bit of ambiguity.

The event was hosted by Rising Minds at the Soho House.

With Product Management Interviews, You Must Listen 90% of the Time


By Dennis Clemente

NEW YORK—At the Product School last August 3, Marty Cagan presented an hour-long talk on product management and the importance of managing new products and features of your startup using a customer delivery toolkit, which includes one important body part – your ear.

“You should be listening 90 percent of the call,” he said, referring to how you let interviewees talk as you hand out questions the rest of it.

But to start, he suggests doing customer research/survey as your build new products:

  1. Understand the market, competitors, user personas and processes
  2. Create a prioritized list of user stories, iterate on designs
  3. Look at impressive existing products

Take these steps, he suggests:

  1. Do a survey. Gather general info about a customer problem and source interviewees, citing how this is fast and scalable (1 hour investment of your time plus get customer response); fun and easy for customers, if you design it right; and everything is recorded and standardized.

The downsides: It’s not good for capturing nuance and details, you will miss unknown unknowns and you can’t drill down into critical areas.

The best practices: keep it short, minimize typing and test on a few internal folks and customers

  1. Do requirements interviews: Get a deeper understanding. Understand user personas, processes

Pros: Captures details and nuances; easier to discover unknown issues/ideas and dig further; develops the customer relationships (for design interviews)

Cons: Takes time (1 hour investment/customer response)

Low sample size (max 5 to 10 interviews per feature)

Best practices:

  • Interview your smartest customers and prospects
  • Send a good pitch email
  • Do a screen share (, hangout)
  • Ask to record it
  • Send questions, document in advance
  • Let the interviewee go off on tangent, if worthwhile
  • Ask if they want to participate interview
  • Once that calls become repetitive, you’re done

Ask good questions

  • Ask broad questions first, then get more narrow
  • Embrace awkward pauses
  • Don’t lead the witness
  • Listen 90 percent of the time

Expand the format

  • Show versus tell
  • Make it interactive

Get some amazing outputs like “a day in the life” description; a flow chart showing current process and variants; prioritizing list of user stories – top ones are your MVP

3. Design interviews. Present design ideas to the user, get input, keep iterating until satisfied

  • Work with a designer to create one or more prototypes
  • Jump on a call and have the user walk through the prototype
  • Iterate based on their feedback
  • When changes start become less frequent, you’re done
  • Can also email prototypes to get feedback more quickly

Best practices

  • Have both PM and designer on the call
  • Get highest fidelity prototype with lowest effort
  • Don’t explain your design, let the user muddle though
  • Don’t argue over design
  • Act on feedback, don’t be stubborn

4.  Do customer visits

Get out and visit customers where they use your product

Pros:  Get the full customer context, tech, processes, personalities, capabilities; gather tons of photos, videos, and competitive research. The downsides? Lots of time (5-hour investment/customer response); can be costly and is a low sample size.

Best practices:

  • Block a few days and bath visits to multiple customers
  • Send agenda in advance
  • Have multiple interview Q&As and/or design prototypes ready
  • If B2B, make sure the right customer stakeholders attend
  • Record as much as possible (with presentation)

5. Do customer summits. Get a group of 5 to 8 sharp customers together and dig into a range of issues.


  • Fun and inspirational for everyone involved
  • Ideas build on each other to unlock new insights
  • Best tool for building relationship


  •  A lot f time (8-hour investment/customer response)
  • A lot of money ($1 K/customer to book venue, possibly flying them in
  • Low sample size

Best practices

  • Easier if customers are local
  • Keep it to 5 to 8 smart representative customers
  • Prepare exercises well (ex. design breakout sessions)
  • Use for brainstorming, requirements gathering, and design
  • Make sure others in their organization get customer exposure

It’s important to use the right tools and activities for the job:

  • Market research (surveys, summits, interviews
  • Requirements fathering (surveys, requirements interviews)
  • UX design (design interviews)
  • Pre-release (alpha/beta testing)
  • Post-release (customer tickets, usability tests, usage data)

Make time for customer research:

  • Worst. You respond ad hoc to your gut-feel ideas, customer request and bug filings
  • Absolute minimum. Do proactive interviews, analyzing usage data and examining customer tickets before deciding what to do
  • Basic. The above, plus some onsite visits quarterly NPS score surveys, and occasional discovery surveys
  • Advanced. The above plus frequent focus groups, on sites and competitive walkthroughs. And you use all of the tools well and adopt the latest tech.

If you forget everything else, here are some more tips:

  1. Gather requirements before design testing
  2. Prepare hard for interviews, mostly listen, and use visuals
  3. Document and share insights with the whole company

Individual customer’s input is one data point. You can choose to ignore it. But when the data lines up, act on it. As for the elephant the room; that is, how do you get people to do the survey with you? Do a good pitch in your email and best of all, give a gift card.

Living in the Time of Audience Rebellion: No One Wants to Pay for Content


By Dennis Clemente

NEW YORK–What’s the future of content? That was the topic at the NY Content Meetup last July 28 at Samsung Accelerator in midtown Manhattan. The title seems ominous — but it has to be. After all, there’s just too much content out there for everyone to consume–and for those in the business of providing content, less and less opportunities to monetize their content. One thing that can save content is good storytelling, despite some ugly realities.

“We’re in the midst of an audience rebellion,” Kirk Cheyfitz said as the moderator of an impressive panel of content marketers. He is the CEO and founder of Story Worldwide, a digital-first ad agency, which has integrated journalistic storytelling, with advertising to tell brands’ stories in every media channel. Cheyfitz has extensive experience in journalism and has started and runs various media agencies.

It’s true. No one wants to pay for content. Not only that, people also want control of what they see and hear. “To force people’s attention is over,” he added. It doesn’t help that people don’t get your content directly to your site anymore. Your content gets linked on, say, Twitter. There are fewer people actually going to sites as social networks have become the default place to go to for newsfeeds, even beyond personal cat postings. And if you think people will click on those banner ads, think again. And we haven’t even mentioned ad blockers.

In a panel of creative people in the content business, there’s always a way.

“Give them some value. Be the destination,” Cheyfitz said.

The youngest in the panel, Bailey Richardson, offered the future as “you being the TV network,” as she also points how Pop-Up Magazine is making journalists, even print journalists not used to the camera or radio, to broadcast the news themselves live. She was one of the early employees at Instagram, where she worked on the community team.

With Pop-Up’s “phone stories” (directly to you), the news becomes very personal, introducing emotion to the (experience).”

Richardson was joined by Jeff Gomez, founder at transmedia Starlight Runner, Mike Knowlton, president and founder at Murmur, a next-generation storytelling studio; Joanne Tombrakos, chief storyteller and business development consultant, and Matt Wellschlager, VP Marketing at Cerosdotcom, an interactive content marketing software company.

The free-flowing conversation resulted in some great insights and new terms on how you can push content that people will listen, read and watch.

  • Take on a micro-narrative approach
  • Where is Harry Potter now? (To stay relevant), it’s now in Lego Games  where it’s the gateway drug for young people go before they go to the (Harry Potter) books and movies
  • Content is the totality, what people are perceiving
  • Work on the Plurals — those who grew completely internet native
  • (Address) the different you in different social networks
  • Do superpositioning. (Example) is Black Lives Matter.
  • Make use of agile /lean development. See if you can fail faster (so you can recover faster)
  • The future of content? Check out why people think mobile wallets is the future.
  • “Listen” to your blind spots
  • Difference of actual story and content? Content is a photo, moments–not always the long narrative arc. Some are primed for content but are not stories

How These Fintech Startups are Changing the Go-To Platform for Investing

By Dennis Clemente

NEW YORK–Last July 27, NewFinance held a Shark Tank-like Open Mike Night featuring financial companies innovating through technology at Rise in Chelsea.

The entrepreneurs were Melanie Smith, founder/CEO at KaChing, a mobile app designed to foster investing habits and financial mindfulness in teenagers;  Rob Koyfman of, a web-based terminal that provides investors with research and visualization tools to analyze stocks, fx, macro and economic indicators and  Marcus Estes, CEO at Chroma, a securities exchange for the private markets economy.

Smith, who taught 9th Grade Math in public schools, said 41 million teens and tweens will earn over $91 billion annually.

“Over the next 30 years 6 trillion will pass to next generation,” she said, who also told about how crucial it may be to add gamification component to her app.

Koyfman, for his part, said he is looking into the 45 million equity investors.

Others included Gregg Jackowitz, founder & CEO at PaidEasy, a mobile payments application that allows users to seamlessly sync with a merchants’ Point of Sales system and João Menano, founder and head of business development at CrowdProcess or James.

PaidEasy is looking solve to app fatigue by having all mobile payments for so many apps integrated on his system, according to Koyfman.

Consumer credit modeling software for risk officers Shuo Zhang, co-founder & CEO at ModernLend, also presented his platform, an online lender that uses alternative data metrics to lend to creditworthy international citizens.

ModernLend is trying to solve the problem for international visitors or students who are in the United States but have no credit score.

“It’s going to be the first credit card for international visitors. You get approved in under 24 hours,” Zhang said.

Last but not the least, Uday Akkaraju, founder at BOND, talked about his intelligent SMS bot that automatically grows users’ wealth based on their financial goals, in the shortest possible duration.

For Bond, why even bother to do this when there are so many platforms out there?

Akkaraju said, Wealth managers are not (always) accessible. BOND is easily accessible

This is very comfortable for everyone,” he said.

The investors at the meetup who asked questions and provided constructive feedback to the entrepreneurs included Matt Perlman, principal at Fenway Summer Ventures, a venture capital firm focused on promising early stage financial services companies;  Alex Tarhini, Associate at Point72 Ventures

Eric Gouin, CEO of Pentalabbs, a Strategic Industrial Investor in Services for Equity that gives to startups power of technology and digital marketing for their acceleration

Meghan Cross, managing mirector of Red Bear Angels, the angel group that invests in early-stage ventures founded or led by Cornell University alumni

Adi Levanon, director of Investments at FinTech Collective, investing in early stage Fintech companies with a focus on banking, lending, payments, capital markets, wealth & asset management, insurance and blockchain based technologies

The meetup was hosted by Ashish Singal.