Today marks a new day for PyCon India. As the whole world already knows, PyCon India has taken a new avatar – A totally online conference. The conference has been a totally enriching experience. They have devised an entirely unique experience, taking the in person format and replicating it to suit the online experience.
There are 4 tech talk tracks (and a inclusion and diversity track ) this year compared to the 3 tracks of previous years. The online format has also brought an influx of international speakers. There is so much going on simultaneously! Since all the talks were online, I was flitting among them and caught of a bit of several talks. I also happened to be traveling today. Yet, I managed to connect. I still miss the ability to connect with people face to face. Given an option I will still choose to attend in person. But given the circumstances, this was the next best alternative.
PyLadies Lunch has been a feature of PyCon India since 2017. Last year, we had further expansion with speed mentoring, quizzes, contests and companies sponsoring PyLadies. More effort was also taken for people of various backgrounds and people with disabilities to actively participate in PyCon India. This year it has been taken much further with a track exclusively dedicated to inclusion and diversity! Hurray! The cherry on top was Naomi Ceder’s Keynote emphasizing the role community plays in times of need to support people who lack the resources and to ensure no one is left behind.
I was a rather passive attendee as I was still on the move during the sessions. The experience on a desktop was excellent though on mobile devices it was not as optimal. I hope to be a bit more participative tomorrow 🙂
Talks I attended
Each and every talk that I sat through was well crafted, coherently presented and informative.
Here are the talks that I attended and some ideas from them that I want to explore.
- Let’s build a no-code tool for small businesses to reduce churn by Padmaja Bhagwat, Bargava Subramanian
They talk about how to evaluate when a customer would churn and how to determine the right time to target them. The interesting part is that they start with data, introduce concepts of customer lifetime (which is the bread and butter of all SaaS companies), progressing to analysis of the data to predict each customer’s lifetime. They finally finish it off which an app built on top of this analysis which other people in the org can use to get insight about each customer.
Some new libraries I learnt
- A comparision of Steamlit with RShiny and Dash : https://plotly.com/comparing-dash-shiny-streamlit/ . Please do note that this article is from Plotly and Dash is also a contribution of Plotly.
- A thousand Djangoes within (or Django multi-tenant) by Lorenzo Peña
I work in Django a lot and so was especially interested in different takes on web related talks.
Lorenzo talks about the different ways multi-tenant systems can be configured in Django. He starts with StackExchange as an example. This is very apt and relatable for almost all listeners.
I am already a bit familiar with multi-tenant systems. One of the key takeaways in this presentation not covered in other PyCon talks on multi-tenant systems are the lesser discussed aspects of multi-tenant databases
- Management Commands
- Celery tasks
- Caching clashes between tenants
- File access restricted by tenant
I hope to go back to this talk to explore these more. My monkey brain was jumping between this and “Taking a peek under the hood – Interpreting black box models”. So I missed some sections of this talk
- Stop Writing Tests! by Zac Hatfield-Dodds
The talk is about Property based testing of code rather than the usual inputs and assert logic. This was something totally new and blew my mind! This is done using the Hypothesis library which was itself inspired from Haskell’s Quickcheck. Love that all the best features of the programming language community are being ported to Python!
More talks that I would like to catch up on when the videos are uploaded
- Towards a more transparent AI – Decrypting ML models using LIME by Laisha Wadhwa
- Taking a peek under the hood – Interpreting black box models by Jennifer Nguyen
- The importance of ‘unconscious bias’ awareness for Diversity and Inclusion
(I promise to update the blog into something more readable tomo.)