I’m Sam and I work as a Data Scientist on the tech team at techspert.io. I use machine learning and data mining techniques to train our AI technology to find the world’s leading specialists to expand our coverage of expert knowledge. Below is an example of my typical workday as a techspertian.
Office visits from furry friends are always a win!
9:00 AM – Time to conquer the day
We have a flexible working policy which gives people the freedom to structure their day around the company’s core hours which are 10 AM - 12 PM and 2 PM - 4 PM. So, as long as you’re in the office during these hours, you can plan your day to suit your schedule. I find I’m more productive in the mornings, this could be due to my daily morning dose of strong caffeine, so I try to arrive by 9 AM on most days. While I wait for the caffeine to fully kick in, I check my emails, Slack (we use this for internal instant messaging) and prepare for the day ahead.
I’m also part of the company’s social committee which means I help organise regular social activities to create that perfect balance of work and play. These socials are a great way for techspertians to spend time together outside of work, get to know each other and have fun. The activities vary and have included punting, games night, go-karting and more. At the moment I’m planning a poker night and need to finalise a few things.
Cruising down the River Cam with some of my colleagues during our punting social. I’m on the left, Ayaka’s sitting next to me and Mike is at the back.
9:30 AM – Begin coding
Now that I’ve got my admin out of the way, it’s time to buckle down and code. I’m currently using natural language processing (NLP) to enable our technology to pick up the commonality between concepts. NLP is a branch of AI that explores how to program computers to process and analyse large amounts of natural language data. It gives our AI technology the power to evaluate how suitable an expert might be for a project, based on content that they’ve published (such as journal articles and research papers). Our back end is almost exclusively written in Python. Due to its simple syntax and a wide variety of libraries specifically aimed at building machine learning models, it’s an effective language to use in AI development.
All our data storage and processing infrastructure is hosted on the AWS cloud platform. I hadn’t used this before joining the team, so it’s been a bit of a steep learning curve, but I’m starting to appreciate just how powerful the service is. More and more companies are moving their infrastructure to the cloud, via platforms like AWS and Microsoft Azure, so this is certainly essential tech knowledge for the future.
It’s time to take a break from coding and catch up with the rest of the tech team.
11:45 AM – Stand up
The tech team meets once a week to provide an update on the previous week and discuss the upcoming one. The conversation focuses on what we’re working on, any challenges we’re facing and our goals. Without catch-ups like this, it’s quite easy to lose track of what’s going on with the rest of the team. I find that articulating what you’re working on and setting goals helps maintain focus and direction and gives everyone a productivity boost. This meeting usually takes about 15 minutes, which is then followed by more coding until the all-important…
1 PM – Lunch
Working in Burleigh Street and near the Grafton, there isn’t a shortage of options when it comes to finding somewhere to grab a bite. Recently I’ve been trying to save damage to my bank balance (and waistline) by bringing my own food, but the company knocked all its goals out the park this month so we’re being treated to pizza on the office balcony (so much for that waistline). Lunch is always a great opportunity to take a break and chat with people which is easy to do here because everyone is friendly and easy to get along with.
Get in my belly!
2 PM – techspertX
Every other week we host techspertX (inspired by TEDx, hence the name). It's a forum to share knowledge, ideas and project updates. Today one of our interns, Preben, presented on the neural network project he’s been working on. Neural networks are computing systems inspired by biological neural systems in living organisms, hence the name. In a manner like that of a human, they attempt to ‘learn’ to perform tasks, given a set of examples. It’s great that we have opportunities like this to learn from each other and share ideas. Preben also wrote a great blog on machine learning vs AI if you want to check it out.
2:30 PM – More coding
Fuelled by pizza and more coffee, it’s time to code some more. This afternoon I’m going to focus on reinforcement learning which is an area of machine learning concerned with how a computer program should take actions in an environment in order to maximise a predefined reward function. I’m using reinforcement learning to improve the accuracy of our web crawlers, thus increasing the number of experts we can draw upon to match with our customers.
I’m all coded out. It’s time to call it a day.
5:30 PM – Home
Home time varies each day, depending on what stage I’m at with my current project, but I’m always glad of the short commute times in Cambridge (the office is a 15-minute cycle from my house). Considering how varied my work is, there really isn’t a typical day at techspert.io, but the above should hopefully give a reasonable representation!
I think I deserve the prize for most enthusiastic techspertian! Also, pizza makes me happy.