Barcamp Saigon 2019 Team
Events, Technology

A Recap of Barcamp Saigon 2019

Here’s a roundup of key discussions that took place during this year’s ‘unconference’.

We just wrapped up one of the most hotly attended tech events in Vietnam, Barcamp Saigon 2019, where over 530 passionate developers in the tech community gathered to network, share and exchange ideas.

We were excited to see a wide range of participants from startup founders, to employees from companies such as Grab, Parcel Perform, Amanotes, KMS Technology, Chotot and more.

Some of Parcel Perform’s team members, including our CTO Khang, and Data Analyst Vy were the main organizers of this year’s event. 

What is Barcamp Saigon?

Barcamp Saigon is an open space, also known as an unconference. While topics are submitted online months before the event, both the content and agenda are determined by participants on the day itself. 

Barcamp Saigon 2019 - choosing topics

Barcamp Saigon - Voting
Participants voting for the topics they’d like to discuss.

Limiting the voting to the event day ensures that the voice of those present is heard. Topics are then allocated to rooms and time slots according to their popularity. 

Below, we recap some of the key sessions discussed during Barcamp Saigon 2019:

1. The “shallow” understanding of deep learning

Currently, the logistics industry still has a long way to go before it gets ‘smart’. Many companies still utilize traditional ways of working: allocate truck drivers to shipment orders on a first-come first-served basis, or select suitable drivers based human knowledge, or at best through basic filtering. 

Thien Pham shared how elements commonly used in face recognition can also be applied to a recommender system, and discussed how deep learning can benefit all the stakeholders involved in the logistics business.

What is deep learning?

Deep learning is a network of artificial neurons with the first layer as input, last layer as output, with many hidden layers in between.

Deep learning

The layers or nodes represent combinations of data points in different ways, allowing the model to reach its given output. For instance, deep learning is applied in the image below:

Deep learning data points

How deep learning works

The model uses a sample of data to assign weight and bias to the nodes in it. By iterating through the data several times (epochs), the model optimizes these weights and biases until the prediction result is stable for a few iterations. New data needs to be continuously tested through the model to determine if the neural network provides accurate information.

In the case of logistics, by analyzing data sets and applying algorithms, supply chain decision-making processes and delivery processes can be optimized significantly. Forecasts of demand and supply can be made too, and with greater accuracy.

2. Common myths about growth hacking

With the boom of tech startups, growth hacking has become one of the top buzzwords in recent years. But can growth truly be “hacked”? Annabel Nguyen from Chotot took the stage to share her thoughts on the buzzword. 

There’s no one-size-fits-all

There are tons of misconceptions about growth hacking, and many think they can simply replicate what has worked for another company and apply it to another. That’s rarely the case.

In order to be an effective growth hacker, you need 3 crucial qualities – creativity, a heart for data, and curiosity. 

You’ll need to think of out-of-the-box strategies to growth hack and help your business acquire and retain customers at a low cost while offering the uniqueness of your company’s product. 

Always test

At the heart of it, growth hacking is to build something quickly to see if it works, and learn if that’s a product your customers want. 

Testing is an iterative process

Keep in mind that experimentation is key. Decide what you want to test while keeping all other variables constant. With so many factors at play in a single situation, it is only through testing that you’ll find what works, and what doesn’t. 

As the needs of your customer evolve, your products need to improve as well to meet their needs. After objectives are set, it should be an iterative process of testing, analyzing and coming up with new ideas to drive growth. 

Growth hacking is an experiment-based marketing informed by data, focusing on product use to drive growth.

3. Using data science for social impact

We live in a world where data is constantly being created. Each day, more than 2.5 quintillion (that’s 2.5 with 18 zeros) bytes of data is produced.

Google search

By the time it took you to read this sentence, over 3.5 billion Google searches have taken place. 

While businesses in the private sector have been building data science capabilities for years, most organizations in the non-profit and public sectors are still far behind. Although they have a strong appetite to use applied data to for their work, they often lack the budget and staff to do so.

When mission-driven organizations have the right talent, tools, and knowledge, data science can generate real human impact.

In the public sector, some ways data science has been applied to solve pressing issues include – having more efficient refugee placement in Switzerland, predicting water demand in drought-stricken California, and even identifying millions around the world who are in need of essentials such as food and healthcare. 

The potential results are clear. By growing the data science capabilities of social and civic organizations, local leaders are able to uncover new insights to build more impactful programs for the communities they serve. 

Our team certainly had a great time discussing meaningful topics that took place during Barcamp Saigon 2019, and we hope you enjoyed yourself too!

If you’re based in Vietnam and looking to join our team of developers to make an impact in the e-commerce logistics industry, check out our job openings here!