Machine Learning is Eating the (Data) World —My DCVC Application

Matt Savio Nicholas
Outlier
Published in
6 min readJan 25, 2020

Note: This update post doubles as my application for Data Collective Venture Capital’s Summer Investment Associate MBA internship in Palo Alto. Check them out here! DCVC is supporting great things in Deep Tech and just closed $725M in fundraising for its 5th round. Fingers crossed that I can help support those great investment initiatives. Also, if you want to apply to the role, fill out their Google Form here! (I don’t mind some healthy competition 😉)

DCVC Request: “Create a Medium post that includes: (1) Favorite xkcd link (2) 150 character max description of a cool machine learning application (3) Paragraph on how you think you can add value in 70 days (4) Optional: additional information you would add in a cover letter”

(1) “Favorite xkcd link”

Hey DCVC! Here’s my favorite xkcd, “CD Tray Fight” (am I dating myself with the CD reference?)

So you needed to check whether the alt-text claim was real, didn’t you?
You can’t read alt-text in Medium, but it states “Randall Munroe unequivocally endorses Matt Savio Nicholas’ views on mindful Machine Learning ”. This may be a rough paraphrase of the original: “This is silly, of course. The enemy will be born in the network.”

(2) “150 character max description of a cool machine learning application”

A haiku competition, is it? Let’s go:

Enterprise Talent

Management! Feed neural net

skill data & outcomes.

Quiz on effective

team composition; output

efficient team builds (150 chars inclusive)

How To Train Your Neutral Net to more effectively staff enterprise employees on cross-functional projects. Note: The 150 character limit applies to both the haiku and parenthetical remark. These haiku(s) are also a small pitch for my co-founded startup, Sift, which (among other things) stores skill data of employees of large enterprises and helps managers build cross-functional project teams using skill matching — I believe the road to automating this work must involve Machine Learning.

(3) “Paragraph on how you think you can add value in 70 days”

Note: Should we assume 70 days is 7 days/week for 10 weeks or 5 days/week for 14 weeks? While I’m available for 14 weeks, let’s assume a 10-week timeframe for now - after all, if I can value in 10 weeks, I can add value in 14 as well:

I see this role as having 3 components: Industry mapping research, portfolio company support, and DCVC business operations. I’m already working on mapping the IoT cybersecurity industry for VC firms with Wharton Venture Foragers. Before arriving, I’ll map DCVC’s cybersecurity investments (like SentinelOne’s IoT “Ranger” product). During my first two weeks, I’ll interview DCVC partners to understand desired research, and read former reports to understanding the format and key deliverables. After aligning on scope, my next 5 weeks will involve researchng that competitive space, identifying early-stage startups, and building the report. After presenting the report, I’ll spend the last 2 weeks building a guide for the next summer associates to get them up to speed. For portfolio company support, I’ll research their business models and press releases before arriving, and pitch ways to support their initiatives (as I did with the CMO of an RPM Ventures portfolio company). The timeline will be similar to the industry mapping, but depends on the portfolio company’s priorities. For DCVC business ops, I’ll interview former DCVC summer associate Adam Brudnick (we know 5 startup-folks in common, including an AI engineer at Samsara and researcher at Quid) to understand DCVC’s current operations. In addition to meeting those needs, I’ll spend 2 weeks mapping ops and 5 weeks developing operations streamlining tools (as I did at Sift and Quicken Loans).

A 10-week timeline outlining Matt Savio Nicholas’ major DCVC initiatives.

(4) “Optional: additional information you would add in a cover letter”

I consider myself a datalogist. As such, I believe that the study of data (the “atoms” of decision-making) is as much a search for fundamental truth as any other science or engineering discipline. I know that machine learning (and specifically “Deep Learning”) has the most immediate potential to radically reshape industries reliant on human pattern recognition. I’m applying to the Summer Investment Associate role at DCVC because I believe in their thesis of Deep Tech transforming global industries, and I believe I have much to offer in support.

Prior Work at a $500M B2B-Tech Venture Capital Fund

While Michigan studying Engineering & Entrepreneurship (with a focus on statistics), I interned at RPM Ventures, a $500M VC fund focused on B2B Tech. During that time, I performed a mix of internal projects for RPM’s managing directors, along with external projects with the CMO of an RPM portfolio company. My experience crafting diligence binders, mapping the IT startup landscape and value chain, and participation in deal pitches, mean that I have the experience to hit the ground running at DCVC. In addition, many of RPM’s portfolio companies, like cybersecurity startup DeepField Networks, ERP software Kontextual, and Autonomous vehicle insurer Avinew were similar to DCVC’s portfolio in Cybersecurity, Enterprise Enablement, and Industry & Transport.

Co-Founder Backed by a Network of Entrepreneurs (and a Presidential Candidate)

After Michigan, I joined Venture for America (VFA), a selective entrepreneurship fellowship led by now-Presidential candidate Andrew Yang. (RPM Ventures’ Managing Director recommended me to Andrew, saying “his passion for startups and technical aptitude […], positive attitude, ability to execute on ideas quickly, and strong work ethic make him an excellent Venture for America candidate.”)

Politics aside, my chats with Andrew have centered around his certainty that innovations in Machine Learning are at the cusp of disrupting human-intensive industries — transportation and manual labor, but also HR, Research, and Management Consulting. These chats informed my career trajectory — I co-founded a startup with other VFA colleagues called Sift, which focused on storing employee data such as skills, interests, and former teams in a searchable directory, making it easier to build teams and manage connections in large enterprises. My colleagues at VFA span the globe today, but are involved in a number of startups and VC firms across industries. We still support one another’s research and career goals, giving me a pre-built startup network.

Experience in Applying Big Data to “Enterprise Enablement”

After growing Sift to over 1.2M in annual revenue, I was offered a job at Quicken Loans to manage large databases for reporting operational improvements. I accepted, because I believe hands-on experience in enterprises would help me understand their needs further. I led a cross-functional team (finance, marketing, IT, and data science) to automate reporting and anomaly detection, eventually developing a dashboard that could propose the most likely causes of declines in profitability before finance analysts begin to look for them.

Current Relevant VC Work Outside my MBA Program

Here at Wharton, I’m studying Business Analytics and Strategic Management. As a member of Wharton Venture Foragers, I’m also compiling a competitive landscape report on cybersecurity for VC firms interested in the future of cyber-physical attacks and endpoint defense. I believe that these go hand-in-hand; as more business functions are integrated with machine/deep learning (through tools like DCVC’s portfolio company Nervana Cloud), more cybersecurity measures must be taken to safeguard business operations.

(5) “P.S. What’s the strangest thing you’re proud about?”

Okay, okay, this wasn’t in the original prompt. As an undergrad, I volunteered to be the guinea pig in my roommate’s thesis that “Anyone can learn to freestyle rap in 6 months (with proper guidance)”. After practicing daily, I freestyled against Erik Torenberg (co-founder of early-stage VC fund Village Global) through his startup rap platform, Rapt.fm. My roommate likes to say he technically “trained a neural net to rap” (bonus xkcd).

“dude could really rap” ~ Erik Torenberg’s review of Matt Savio Nicholas

(6) Deadlines to Note: February 24th, 2020

--

--

Matt Savio Nicholas
Outlier

I use technology to understand humanity. @Venture4America fellow, @UMich alum, launching something new in Detroit. ⚜