I am amazed that content of this quality was presented by community members for free. It was totally worth being inside and roaming the maze-like halls of Best Buy Corporate Headquarters on one of the few sunny Saturdays we get up here in the frozen tundra of Minnesota. The exposure to new technical topics was great, but more importantly experiencing the energy of the people who are active in the Minnesota tech community was the real core of the experience.
I will try to mirror the energy and great themes of each of the sessions that I attended. The keyword is ‘try’. I apologize in advance if the energy and competence of each session I attended doesn’t shine through. Hopefully, some of the links I am adding to each session will help you navigate to additional resources.
This session consisted of a panel of people active in the Minnesota startup community.
- Jeff Pesek – Moderator from tech.mn a news site dedicated to the startup community within Minnesota.
- Glafira Marcon – Lead organizer of Healthcare.mn
- Dan Atkins – Co-founder of MinneAnalytics
- Ryan Broshar – Techstars and Matchstick Ventures
- Justin Grammens – Co-founder of IoTFuse
- Kevin Spanbauer – Dedicated to finding and providing resources and people to software startups within Minnesota
The general takeaways that I got from the panel were:
- Get involved with the community
- Don’t be afraid to share your ideas. Sharing your idea will only make it stronger.
- Double down when it seems like you should quit, double down again when it seems like you should stop, and when it gets hard and you think you are truly at the end, double down again
- There is help for anything you want to do with your startup. From business, to community, to technology, there is an amazing community here in Minnesota that is dedicated to helping you get your business started in some way.
- In general the panel is encouraged at the state of the startup and general business climate in Minnesota.
- The depth of participation in the MN Cup Startup Competition was cited as a great sign of the vitality of the startup community within Minnesota.
The process and overall maturity of the way in which the U of M incubates and manages the movement of their technology portfolio from the university into the business sector is astounding.
I can’t do the whole presentation justice, but for future reference you should watch the U of M Entrepreneurs site. Especially great is the University Startup Pipeline Google Doc which shows the stages of the startups as they move from the university onto business.
Ben Peter gave a great presentation on how to use the Zapier service to create custom data workflows from existing third party Web APIs.
His best demo was a code-free custom workflow that would send a free t-shirt to an attendee given an e-mail address contained in a text message. The demo used Zapier to aggregate SMS, Google Sheets, and Printfection, all without code to create a custom business promotion on-the-fly.
There is real power in being able to aggregate the features of entire businesses via their API surface. This power allows startups to create easy solutions for repetitive business tasks in a fraction of the time and resources of having to write custom code or rev up whole hosts and custom service silos.
Andrew Rahn of Iconfactory gave a great side by side talk that compared Kotlin and Swift.
I really enjoyed his openness, energy, and passion for this topic. His openness really made the talk a true two-way discussion between himself and the audience.
Personally, I have done some Swift, and absolutely no Kotlin. His session served as a great jump start for me as I learn more about the Kotlin language.
The general take aways were:
- Both Swift and Kotlin are really new languages, and as such are in a state of flux as their owners tweak the compilers, syntax, and environments
- Both Swift and Kotlin are modern languages that force the developer to use the compiler to check nullability and mutability rules
- Swift and Kotlin suffer greatly from being bolted on top of their respective legacy heritage (Swift — CocoaTouch / Objective-C / Automatic Reference Counting — Kotlin – Java / Android)
- I would be incredibly lucky to ever get a chance to work with Andrew Rahn on any project.
I went out of my comfort zone of front-end UI development and attended this panel.
Please see the minnebar session page for the panel members and full description of the session.
- When it comes to data science you spend over 90% of your time cleaning up your data sets before you can actually do any analysis. Things like different date formats, missing or null values, punctuation, and differing data entry techniques cause the data scientist to spend a ton of time cleaning up data.
- There is a debate within the data science community between those who are going hunting through data looking for patterns, and more hard science statisticians who seek to apply statistical techniques to data to achieve a less biased / less ‘hunt and peck’ methodology to looking at mass data sets.
- The Analyze This! competition seems like a really great thing to take a look at. The Analyze This competition is a long term (3+ months) competition that goes from raw data through to creating a possible predictive model.
- Another interesting resource that was mentioned was Kaggle. Kaggle is a source of competitions and sample datasets.
At heart I am a native front-end developer, and this talk was 100% in my wheelhouse.
Their work brings the architectural ideas of React and the concepts in Flux to front-end iOS development using Swift.
Their presentation was done so well that the core concepts they presented could be applied to clean up and simplify almost any front-end development you may be doing on any platform.
Their work has inspired me to start pursuing a similar framework for Xamarin.iOS and Mono.Android development. In fact, the more I think about, the more I realize that their work may be more powerful when expressed in a cross-platform framework as it could lead to incredibly thick UI code sharing between platforms, and incredibly thin per-platform code. The possibilities for eliminating MVCs (Massive View Controllers) or huge overwrought Android Activity code are awesome.
If you do any UI work at all on any platform: Run, don’t walk, to their presentation slides and see the future.
Justin had some great examples of applying the Google Prediction API against the Nice Ride data set to try and predict where a bike would be dropped off depending on when and where it was picked up.
The main question from the audience during Justin’s demos was: What is the underlying statistical model that the Google Prediction API uses? The answer, unfortunately, is that Google isn’t telling.
I would file the Google Prediction API away as a future tool for your toolbelt. At least until Google takes it away, never to be seen again.