What organizing tech talks at the office taught me

Around June 2014, I started organizing technical talks at my office. The organization I work for is not small and we do have a technical council that organizes trainings and an annual coding competition. The quality of many of the trainings are good. I personally did a series of trainings on advanced algorithms and data structures by Dr. Naveen Garg and it was very very good.

But, I realized we were missing something. The trainings planned by a central team was top-down. It should also be supplemented with bottom-up employee driven informal trainings and technical talks. We write verification tools for the semiconductor industries. Domain knowledge is a very important part of our job. But we essentially build software. If the foundation of software is weak, then in my opinion, even a brilliant domain specific idea can not produce a better product which can be reliably used by our customers. And towards that goal, I thought we should organize more informal tech-talks.

I approached my manager and his manager and both of them were interested. I started off with a video screening of “Insecure coding in C and C++” by Olve Maudal. I thought it would be a super hit since I loved watching it. But it was not! People preferred a more interactive technical session.

So far, in the last six months I organized 5 tech-talks which is far less than what I thought I would do. But I am glad that I got at least 4 people agreeing to present on different topics. What I learned from organizing the tech-talks are –
1. Organizing them on your own takes a little effort but it is definitely a good experience.
2. A technical talk should not exceed 1 hour. People do not have more than one hour to spend on a topic which may not align with their work. And there should be at least 15 minutes of discussion.
3. A technical talk should never happen on Friday post-lunch as people tend to leave early on that day.
4. A technical talk should not focus only on the domain in which we work, but also on other domains too. We had a technical talk on “k-means based document clustering” and it gave us a few ideas which can be applied to our work.