I’m pleased to run the sixth interview in our #testerstory series with our guest Laveena Ramchandani from London, UK. She will tell us about her journey, how she combined Data science and testing and the perspectives that could be generated. 

Hope it will inspire lots of you who are looking for Data Science and testing careers oriented.

Part 1: Introduction and work from home experience

1. Tell us about your experience, background? 

I am a British born and brought up in the Canary Islands. I studied at a British school therefore decided to do my undergraduate studies in the UK. The course I studied was Computer science and I thoroughly enjoyed that stage of my life. I enjoy testing as well as baking and yoga.

2. How did you join the world of testing? 

My testing adventure started from a graduate role. I fell into testing like most fellow testers I hear. Testing was not covered in depth at university, so It was really new for me. I learnt everything on the job from day one and today itā€™s been over seven years that I have been working and serving the community as a tester. I have had experience in functional and non-functional testing. I enjoy my job and would love to learn more and share it too.

3. Tell us more about Data Science, BI and other functional domains that you worked in ?

I had no idea the project I would be working as a tester would be data science based. Thanks to a lovely team and helpful members with whom I learnt so much and have so much to share about it too. It has definitely been an interesting area to work in and I appreciate this is quite new in the testing community. I hear many testers know of data science teams and the product jointly being built but are not directly working with the data science team. I think itā€™s a great opportunity for testers to start getting involved in.

My first ever role was as a BI tester, and it was quite interesting as I was working for one of the renowned oil & gas firms and analysing their BI reports. I learnt about ETL in my first ever role which helped me a lot in my second job at TFL too.

In a nut-shell I have worked with a lot of data and would always suggest it’s vital to understand the data you are working with and if any data is being edited then make sure you are aware too before it generates biased results!

4. Letā€™s talk about 2020 labelled as the remote year, tell us about your move to remote agile mode?

2020 has been a difficult situation for many and I empathise with it. It was not easy to work from home and not see your colleagues or have team lunches. It was for sure a different way to work and now slowly I have gotten used to it. However, 2020 has also been a magical year for me in terms of my career. It was the first year I decided to go out there in the community and share my knowledge. I spoke at 10 international conferences, got interviewed for 4 different podcasts, 5 written interviews, 4 blog collaborations, wrote about testing in ā€œaround the world with 80 software testersā€ by Viv Richards, and 5 blogs independent blogs. It has been such a great experience and I cannot wait for 2021 to come up with more experiences like this.

5.   How did you face agile practices remotely? How do you maintain the same quality of work when everyone is working from home?

I think it was important to stay in touch with the team a lot more. Make sure to ask many questions and have a better understanding as online meetings can sometimes get disrupted. I think the main message Iā€™d like to give here is to feel like a team, make sure you put in all the effort and make yourself available for the team. I saw myself asking plenty more questions as it was allowing me to design my tests better. As for the team I was working in, we all would agree and raise things in our retroā€™s to improve not just our product but the team practises too.

Part 2: Data Science and Testing

6. Testing is context dependent, Being a tester you can work in many fields. From a tester vision, what is the consequence and the impact on your career when moving from field to another? 

I agree it can be a bit overwhelming as it can be a lot to absorb when moving from one field to another. However, I always think if one desires to move fields they should as a particular field may not be a perfect match for someone whereas it could be, and excellent match for another person. I personally look at fields as a puzzle if you donā€™t fit in one part you will definitely fit elsewhere. The world of IT is ever growing and there are so many opportunities so why not explore?

As per myself, I tend to shift a little left, as a tester I perform my testing role, but I also try to learn around product design UX or look into DevOps. I like to spread my wings.

7. I see that you have considerable experience in data science, Tell us more about the role of a Data Scientist. Is It mandatory to go deep into the technical side (automation/algorithms)?

Data science is a comprehensive process that involves pre-processing analysis, visualization and prediction. It comprises of statistical techniques and it is more of data driven decisions. A data science model basically provides a mechanism to optimise results.

As a tester I bought my suitcase full of testing knowledge and skill set, which was quite handy. But as with any new project, new areas exist. So, I learnt how the model functions and what type of algorithms we use. In terms of data, I tried to understand how this data is adding value via the model to the client and whether the results are adding value for the client.

It is not mandatory to have a data science background, I would suggest doing some research and understanding from your data scientists how they are creating new features and what algorithms they are using.

Some of the useful algorithms I have worked with are Genetic algorithms, Monte Carlo simulation, Brent’s method and some business logic rules.

8. I suppose there are many tools that you use to maintain your tests while testing data models. What kind of tools are recommended and how do you choose them?

I have worked with both background testing and frontend testing. I would suggest if you are working with databases then have some skills around SQL queries and statistics. In terms of frontend, I would suggest using cypress testing. Also, a testing pyramid can also be a good option to bring into your teams when working on testing strategies.

9. Is it better to start in Data Science then add testing or from testing then add Data Science?

There is no hard and fast rule on whatā€™s better. My suggestion will be whichever way you start, make sure you take all your testing knowledge and learn the new skills too. Be like a sponge, take in as much useful knowledge as possible as it can help create better tests and add quality.

It is OK not to know everything too, there are SO many algorithms as it is! But having a basic understanding is enough.

10.  Both Testing and Data Science are considered as brilliant careers paths to consider for IT people. How did you combine them? 

As I mentioned above, I took my testing skills with me and used them as well as learnt how to test a data science model. I would suggest the same to everyone.

11. In what parts Data Science can help your testing?  

Some potential risks are around having bad data, bad analytics, data costs, privacy, security and the selection of a wrong algorithm. Therefore, understanding the risks helped me test the product better. If the data is not good, then the validation will be wrong. Also, data storage costs money, so itā€™s best to plan well and have data to the right granularity to avoid spiralling costs. In terms of privacy and security be careful and anonymise any client data where possible. Selection of a wrong algorithm might not give the best results and may lead to bad model validation.

Make sure to have a risks checklist as that will help improve areas on where testing may need to be more critical.

12. How do you test a complex Data Science Model?  Tell us about algorithms or models you use to drive your testing (do you have an example that could explain the relation between testing and data science?)

I tested it like any other product. Back end testing involved running procedures and making sure the database is up to date and I can query it and validate my results. In terms of front end any new functionality or new client data impacts were testing. I noticed that the data model I work with provides insightful information to its clients, but in some areas, it is stochastic(random). This is what makes it special too; the results can deviate a little bit but having a threshold in place like “we can accept changes between 3-5%” is something new that I learned. It was crucial to also learn from a data scientist how I know my results are accurate or not.

As mentioned above, I learned that having thresholds was a good option for a model that delivers results to optimise a clientā€™s requests. If a model is stochastic then that means certain parts will have results which may look wrong, but they are actually not. For instance, 5 + 3 = 8 for all of us but the model may output 5.0003, which is not wrong, but with a stochastic model what was useful was adding thresholds of what we could and couldnā€™t accept. I would definitely recommend trying to add thresholds; if your model is providing important insights around fraud detection, for instance, then your threshold should be even lower, around 0.5-1% acceptance of deviation in results.

Make sure you also understand your data set, as well as all the configurations that are being inserted to make the model run and provide optimised results.

I have learned a lot about how results can be stochastic and how we can accept results within a certain threshold.

Also, configurations to run a model are something new I had learned, so it was vital to understand what each of them meant to the model and client. Each client would require different information for each configuration. It was useful to do some exploratory testing on the configurations and see what the model was providing me with.

Part 3: Conclusion 

13.   What advice do you give for both junior or senior testers who want to move to other functional fields such as Data Science or other .. ?

Making sure we have enough information about a new client requirement and the team understands it

Validating results including results which are stochastic

Making sure the results make sense

Making sure the product does not break

Making sure no repetitive bugs are found, in other words, a bug has been fixed properly

Making sure to pair up with developers and data scientists to understand a feature better

If you have a front end dashboard showcasing your results, making sure the details all make sense and have done some accessibility testing on it too

Testing the performance of the runs as well, if they take longer due to certain configurations or not

Books ā€“ accelerating software quality: Machine learning by Eran Kinsbruner

Donā€™t hesitate to follow me on Twitter and Linkedin Laveena Ramchandani

15. Anything else you want to share with the testing community ?

Testing is a very interesting field, there is a lot to learn here and share. There is a huge community out there too which is excellent as everyone is super helpful and insightful. I would suggest attending some webinars and conferences and maybe one day even try to be a speaker. It is greatly appreciated too if you can share knowledge towards the community even via blogs. 

Testing is a mindset independent in which field you are working. 

But it’s better that you choose a field that you are passionate about it. 

So, combination of double passion can lead to magic and facilitate career growth in both directions.

Thank you Laveena for being part of this interview, sharing your unique experience and thoughts about both software testing and Data Science. You are a leader to follow in the testing community. 

I encourage you to continue on your brilliant journey, wish you all the best in your career. 

If you want to discover the latest testing trends, check Laveenaā€™s article about testing a Data Science model or check out deeper technical blogs on her medium.

Thank you Laveena for being part of this interview, sharing your unique experience and thoughts about both software testing and Data Science. You are a leader to follow in the testing community. 

I encourage you to continue on your brilliant journey, wish you all the best in your career. 

If you want to discover the latest testing trends, check Laveenaā€™s article about testing a Data Science model or check out deeper technical blogs on her medium.

Interviews History