About Kevin Marshall: More than you ever wanted to know about my career path.

I recently had a college student (and one of my son’s roommates) ask to interview me about my career path for one of his CS class requirements...and as it turns out, I have a really old blog post that fleshed most of that out (in way more detail than anyone would probably ever want)...so I dusted that off, updated it a bit to catch up with today...and am sharing it here now (mostly just because I did the work of dusting it off, and now want to publish it *somewhere*).

The meat of this post was originally inspired by the publishing of the USV Timeline.

What follows is a rambling bullet format of that:

1992

Graduated Girard high school

Attended Edinboro University (initial major - Secondary Mathmatics, eventually switched to Computer Science Major but never achieve degree. Fun fact, the school doesn’t even exist anymore. It’s now part of Penn West.)

1993/1994

Picked up HTML from early issues of Net Magazine; built personal website hosted on the Edinboro VAX system.

Hired at $5/hr to 'build websites' for local video production company (Pro Video Productions).

Taught myself Perl (thanks in large part to the O'Reilly 'the pink camel' book). Without any knowledge of databases, built listing service for local RE/MAX office (including very rough, but workable, CMS and search feature [taking heavy advantage of Perl regular expressions]).

1995/1997

Began to do a variety of freelance web development and programming work (heavy use of Perl, HTML, and Photoshop in these days).

Created a CD-ROM project in association with a local ad agency for Erie Institute of Technology (built with Shockwave)

Created a nationally distributed CD-ROM for Truck Lite sales staff (programmed in Visual Basic).

Took a full time position with Corry Publishing to be their 'web' guy (also did a large amount of networking and hardware support while there). Introduced to Coldfusion and began working heavily with Microsoft Access (and eventually MS SQL). Initial salary was $24k/year; ending salary was, I think, around $32k/year

Introduced to the concept of fantasy football (won the league the 1st year, 2nd place the 2nd -; hooked). Begin to build web tools to help manage my team and our league.

1997-1998

Friend recruits me to American Express in Greensboro, North Carolina. Primarily work in ASP and MS SQL. Initial salary of $40k year (and chance to focus on just software/programming). The move to NC officially ends my college efforts (until this point I had continued to take one or two classes a semester; but had serious doubts about the usefulness of my major and was considering moving to a 'business' major because I thought it might be more complimentary to my now growing real-world experience).

Create tracking system estimated to save Amex over a million a year.

Begin to expand fantasy football software (thedfl.com at the time) to allow other leagues/users.

1998-1999

During weekend trip to NYC, meet my wife. Quickly decide to make the move. Do another weekend of 9 interviews, get 8 offers. Pick <KPE> because of the variety of projects the advertising firm connections brings, also buy into the environment and the possibility of incubating some of my own ideas there. Initial salary $90k (wasn't highest offer on the table)

Move mostly back to client work with Coldfusion (doing a little ASP project work here and there). Heavy MS SQL for the backend and begin to dabble in Unix/Linux for server hosting. Also begin to dabble in Java.

Create and build BML (Barney Markup Language) as key part to the CMS system for the e-specially my Barney project (1st physical product I worked on and that was available in toy stores nation wide).

Pitch <KPE> on locagame.com idea (rate parks for pickup games and facilities/activities; run tournaments; etc.). Begin to have concept developed (eventually killed by higher ups for reasons never fully explained to me).

Slightly disgruntled by the LocalGame experience and also concerned about lack of profitable projects coming in (compared to cost of parties and events being hosted) I change jobs moving to reviews.com. Finally break the $100k/year salary range (having accomplished an early personal goal of earning a 100k salary, I no longer focus on salary when evaluating opportunities).

Ideas and concepts for websites, along with hacking out initial prototypes and services begin to be a passion/hobby for me. Finally have the personal funds to start buying domain names for all of these ideas (will eventually cycle through at least 200+ domains and ideas; some still remain in my collection, many do not.)

1999-2003

Armed with screen shots and data feed from the ACM, build small team and version one of reviews.com (Colfusion, Java, and Oracle).

Expand thedfl.com into Supermug.com (pitch as software/service to run fantasy football leagues for bars and promote specials to customers throughout the week; fail to gain traction). Never makes money.

Break out draft tools and advice within Supermug to it's own product Draftwizard.com (fantasy football help site/forum; subscription product; gain small amount of traction but break-even financial at best -; make many great friends from all around the country via the service; become 'fantasy football expert' in variety of national fantasy football magazines). Invest about $40k over the years; Basically break even.

Build statsfeed.com service to help power Supermug and Draftwizard. Connect with other small fantasy services/sites via the Fantasy Sports Trade Association (FSTA) and get handful of b2b subscribers. Becomes solid lifestyle business during football seasons. Without realizing it, build one of the 1st b2b webservices of the time serving up XML on demand (primarily Java and MS SQL backend). Earns small 5-6 figure profit for a couple of years (mostly used to fund my other projects).

Begin to do a small amount of writing (primarily reviews for the reviews.com product; some articles for fantasy football publications and my advertising).

2004-2008

Thanks in part to a variety of factors, decide to take time off to stay home with my 1st born son (born Oct. 2003). Agree to remain as part-time consultant with reviews.com until they no longer need me (end up doing various small tidbits for them up through the mid 2010s).

While being a stay-at-home dad, give up on Supermug concept and stop investing in Draftwizard concept (maintain for a few years before completely shutting it down). Continue to keep Statsfeed as a small, niche business but lack interest to grow it (eventually shut it down as well to focus on other things).

Frustrated with the effort involved in Java, the work and cost involved in trying to scale Coldfusion, begin to port all my personal projects and legacy systems to Ruby on Rails.

Mostly because of recommendation/introduction via the DBA I hired while at reviews.com (who went on to write the SQL Cookbook for O'Reilly among other great things), begin to contribute to some O'Reilly books. Eventually get deal to write 'Web Services with Rails' PDF for O'Reilly. This leads to a book deal with Apress to write 'Pro ActiveRecord: Databases with Ruby and Rails'. Very little profit in writing, but earns me a lot of 'street credit', job offers, and better consulting rates.

Pitch book idea for 'Adding search to your site' to O'Reilly and get approval to write explore the idea a bit further (get side tracked on other projects and never end up pushing the idea forward).

Get approval from Apress to write a book on Amazon SimpleDB (write 90% of the book, but fall way behind on schedule and end up scrapping the project before publication [Apress was going through a large re-org at the time and since we had never done an advance on the project it was easy for all parties to just scrap once interest/passion was lost -; and it became apparent SimpleDB was not going to be as big a hit as we initially thought]).

2008-2009

Decide I'm ready to get back into more 'full time' work. Recruiter places me at Bowker on a 1 year contract (and I pick up my activity with reviews.com again). After the initial year with Bowker, offered a full time position and stay for a couple more months before deciding 'corporate' (and layers of management) are just not my thing any more (final straw comes when I get reprimanded for agreeing to give a $20 refund to a customer that encountered a bug in a service I wrote [doing what’s right for the customer is a very strong, personal, core value of mine]).

Upon request go back to reviews.com full time...with the communication/intent that it will be short term while I figure out what I really want to do long term.

Begin to heavily read blogs and learn about the NYC tech. scene that had developed over the years under my nose (without my knowledge or involvement). Connect with Charlie O'Donnell just as he's starting Path101 with Alex Lines -; these two become the root of my entire 'tech' network and become directly responsible for 90+ percent of the connections I've made since.

Having moved away from the sports and fantasy businesses, begin to take a heavy interest in social data and public/free APIs. Build out many quick hacks and prototypes of various 'fun' ideas (some with friends and connections, some as personal projects). All with an eye towards 'starting something'. Start doing a large amount of PHP and have now moved most of my personal projects into a PostgreSQL backend (cheapest hosting options, with most flexibility, for my quick hacks). conversationlist.com is probably the most well-known of these, but many find various levels of limited success (for example AppsiGot, literally a 24hr hack that I never touch again, wins a runner-up status in the 1st Techcrunch NYC hackathon [GroupMe was the big winner at that hackathon]).

Via an introduction from Charlie, decide to join a small startup called Catchafire.org and help build out the tech. Enjoy the team, but not really passionate about the concept (hindsight, I made the jump more to jump than anything else). Believing my lack of passion will be a huge long-term challenge for the team, I quickly transition out.

Via an introduction from Terry Jones I meet John Borthwick and am recruited to work with bit.ly (mostly jump just at the opportunity to work directly with Hilary Mason; also like the idea of potential to work with Alex Lines; and fascinated from a distance by the BetaWorks model and the scale/data of bit.ly). I accept the deal as a consultant making it clear that I will eventually leave to do my own startup (though I didn't yet know what that startup would be). While there I learn a ton about scaling, move heavily into Python, and fall deep in love with MongoDB. I build out the prototype/v1 of the bundles API.

Via an ad-agency connection (and friend) I take a contract gig to build out a BlackBerry, Android, and iPhone app for Novartis (the profit from this project ends up being used to fund a year of KnowAbout.it efforts).

2010-2011

Will Cole (another Charlie O. connection) and I decide to make run at turning KnowAbout.it into a real business. Because of growing concern that news.me (a bit.ly owned/operated product) will eventually pivot into what I was already doing with KnowAbout.it I decide to end my contract with bit.ly (attempted to leave in Nov. but was convinced to stay until Jan). This essentially puts me full time into KnowAbout.it for the next year.

I build out a handful of, non-critical to KnowAbout.it, support services. These include click tracking with halfbite.com and content-parsing with uridata.com (since killed).

Initially built as a social data search tool, KnowAbout.it evolved into a social discovery tool (it used the data you created across social networks to figure out, and perform on your behalf, searches for things being shared with you that you likely wanted to know about). System scaled well and processed/indexed millions of links a day...and generated an average of a 40% click-through rate for those that opened the daily emails...but ultimately failed to gain enough users, quick enough, to be considered a success.

Failing to gain enough traction with KnowAbout.it to raise funding, and at the urging of a recruiter, I pick up a gig with MyYearbook (hoping to replenish the bank account a bit). Quickly regret the decision (love the team, hate the product, so-so on the actual tech, re-discover that I'm not a good fit for the stage of the company [they were transitioning from small to medium; and had just sold for 100 million plus]). Express my concerns to the management and we agree (I think on good terms) to part ways.

Finally build out Android and iPhone versions of 'Report to Santa' app I've had bouncing around in my head for a couple of years (don't push it enough or try to monetize it but it gets downloaded by over 10,000 people).

Via an introduction from Darren Herman, meet with founders of Pubgears. Love the history, love the guys, love that it's profitable, love that it's a reason to get out of my gig with MyYearbook. Join the team with the intent of building out the tech. and help the company scale.

Begin acquisition talks with a couple of interested parties around Knowabout.it. Fail to work out a deal (primarily because I'm already committed to PubGears and unwilling to be including as part of the acquisition). Decide to close the service down (to save on out-of-pocket server fees).

2012-2016

In an attempt to cut down on server and domain costs (as well as narrow my focus) I begin to quietly shut down a large variety of my hacks and little projects that have no real traction (there are over 100 with some level/version of code in existence over the years by this point in time).

Focusing the majority of my time and energy into building out tech for PubGears, I find myself falling out of touch with sports news...so I build turfd.com (a system that crawls and indexes sports news from around the web; I then build a feature to automatically search that index for my favorite teams and send me the new results via email).

Sparked by a handful of offline conversations, I decide that the Turfd system could be expanded into a number of interesting use-cases...and because of my interest/personal history with the AVC community, I decide that applying it to the discussions there might be an interesting start (I label the service Gawk.it and turn it into a search engine for blogs that allows users to search the blog posts and all the blog comments). I build out the initial version over a weekend, play with it a bit, and then decide to point it out to Fred Wilson. He bounces back some initial feedback and ideas that I iterate on, and then decides to install it on avc.com and gothamgal.com (as well as blog about it). I spend the rest of the summer days pushing PubGears tech forward as much as possible and the nights and weekends improving and pushing the Gawk.it product/idea forward.

After some good initial traction, Gawk.it stalls out around 100 or so high profile blog installs (blogs are becoming less and less popular as social media is taking over the content world). I eventually decide to direct all the Gawk.it traffic to duckduckgo.com and focus on other side projects with potential.

Alexa is launched and provides an interesting side project distraction. I build one of the very early Alexa skills and call it Math Mania which picks up some early attention (and to this day continues to earn me small bits of AWS credits and pay outs). Through the years I'll go in-and-out of building various voice-based Alexa games & skills (with various levels of traction and rewards).

PubGears itself also hits a bit of a plateau around the 10 million revenue mark and we all begin to think about what might be next. Ultimately we decide it's time to sell the company (it eventually gets acquired by NextTag).

Rather than being a part of the acquisition, I decide to look into starting the next thing. Lance Lovette (who I originally met during that brief Catchafire stint) connects me with a couple of his old peers from his Optimost days, Scott Simonelli and Andrew Eisner, who have just launched the idea of "the first audio analytics platform" (Veritonic) and are looking for a technical co-founder to help make the vision a reality. Once again, I love the team, love the vision, and love the potential for some interesting technical challenges -- I decide to join.

2016-2022

I begin to focus the majority of my time and energy into building out tech for Veritonic (along with the help from some key advisors, I build out some of the first Machine Learning models that work directly with raw audio along with the rest of the platform, tools, and team).

Crypto breaks onto the scene and I spend a lot of side project time digging into the opportunities (eventually getting on the radar with a project called TokenTogether, I end up doing some small freelance work for Coinbase). However, my commitment to Veritonic is requiring too much time for me to focus too much on the crypto space so ultimately I do very little with it.

While at a game night with some friends, I end up with the idea to challenge myself to develop my own table top game. Eventually I come up with the idea for FubNub & Bad, Bad, Turtle...which I successful run a Kickstarter campaign to launch. As I dig deeper into the table top world, learning a lot about the space and the opportunities, I end up developing two more games (Bodimeca and the Four bears AND Herobrawl and the Roshambo Protocol). I learn how to produce and ship the games in low volume and affordably (allowing Hero Brawl to show small but successful early growth).

Interested in the micro-saas movement, I also spin up the TextWorkouts idea. It shows early success, but is not a vertical I'm personally passionate about, so I find a partner to sell the service too (he’s planning to formally relaunch it soon).

As of the time of writing this, I continue to focus most of my time on building and growing the Veritonic tech (and tech team). Veritonic itself is now the leading audio analytics platform and scaling quickly (yeah! We actually are doing it!).

I also continue hacking away on various small, focused, side projects that help me continue to learn a new technology, business concept, or expand my skills or knowledge some other interesting way.