Spotlight on Women in Tech: Dr Sarvnaz Karimi
A Principal Research Scientist at CSIRO’s Data61, Dr Sarvnaz Karimi has over 12 years of experience working in health informatics, applying her expertise at the intersection of Information Retrieval, Natural Language Processing (NLP), and Machine Learning.
Here we chat to her about what inspired her to start a career in tech, decision support systems for clinicians, how she’s applying NLP to identify outbreaks, and her advice for anyone looking to enter the industry.
What led you to choose a career in tech?
I enrolled in a course called ‘Computers’ at my high school that introduced computer science and programming. I found coding fascinating. You could build something by carefully putting together some commands. It tapped into my problem-solving skills; it was intriguing.
At university, I studied software engineering (graduating with a university medal), followed by a masters in artificial intelligence (AI) and then a PhD in natural language processing (NLP). The combination of needing to be creative and dynamic to solve problems and my ambition to constantly learn were my drivers.
What attracted you to join CSIRO’s Data61?
CSIRO is a well-established organisation known for its world-class interdisciplinary science. When I saw a job opening in NLP in 2012, I thought this could be my chance to become a scientist who works on real-world problems in collaboration with other scientists from different fields.
Can you please describe your professional background and the areas you specialise in?
After finishing my PhD in 2008, I joined NICTA (National ICT Australia) and the University of Melbourne as a postdoctoral researcher to work on BioTALA (Biomedical Text and Language Applications).
Four years later, I joined the NLP team at CSIRO’s ICT centre as a research scientist, and 10 years later I’m now a principal research scientist at CSIRO’s Data61.I’m also the Deputy Director of Publicity for Association for Computational Linguistics (ACL) and a former President of Australasian Language Technology Association (ALTA).
I specialised in NLP and Search and Information Retrieval (IR), specifically in applying NLP and IR techniques to different domain applications and science fields where textual data is available. I’m particularly interested in a sub-field of NLP known as ‘information extraction.’ It allows you to automatically extract information from text to reveal new insights and better assist decision making.
Can you share an example of a data and digital science project that you’ve worked on that you’re most proud of, and that has achieved positive impact? What was the biggest lesson you took away from the experience?
One of my most recent projects is the development of a decision-making system for doctors that can save vast amounts of time and help provide better medical care for their patients. It integrates the medical notes and history of patients before matching them to clinical trials and scientific literature that could assist their clinicians to diagnose or treat them.This potentially cuts months of searching for suitable clinical trials for patients with cancer, blood disorders, etc.
My team and I are also working on another medical decision support system designed to help GPs ensure their patients are healthier for longer. The system integrates both individual and family medical history to automatically remind GPs when the patient needs a crucial test that could predict or prevent future disease. These reminders are useful if the patient’s regular GP is unavailable or exceptionally busy.
I’m also working on an NLP-driven epidemic intelligence decision support system that can identify potential outbreaks of infectious diseases before they publicly arise.
It identifies key words and phrases from numerous online sources, including formal publications like technical articles, to informal channels like social media. The system can be adapted for multiple illness, including Covid-19, influenza and gastro outbreaks. It can also track current infections and predict future spread.
One lesson I’ve learned from past projects is to never assume the same method will work everywhere. New data and problems bring their own subtle differences that can lead us to use and develop yet another technique.
What are some of the projects you’re working on at Data61? What about them excites you?
I’ve been applying text mining methods to precision health and social media analysis to assist in human decision making. NLP can accelerate many of the processes that have been traditionally done manually and require considerable human effort. We can save time and lives by helping people do their job faster and sometimes even more accurately.
In your opinion, what’s the single biggest change that needs to happen to encourage more women to pursue careers in tech?
Removing stereotypes from early childhood. It has to change among parents and educators. Even when we want to address this, we should use the right questions to avoid planting the idea of what is right or wrong for any gender to be in any profession.
How can colleagues, organisations and industries within tech better support and enable women?
Having women in management positions, promoting a culture of inclusiveness (gender, ethnicity, and all other aspects of diversity) and providing channels to report wrong behaviour while feeling safe.
What advice would you give to women and girls wanting to pursue a career in tech?
It’s one of the most empowering and rewarding careers one could pursue. More and more workplaces are going to need people with technical skills (such as programming or a working knowledge of machine learning) and you’ll never run out of options.It’s also one of the jobs that can be done from the comfort of the home providing even more flexibility if needed.
It’s easy to be intimidated by some of the existing cultures in the tech world, but once you get past that, it can be extremely rewarding.
Discover some of Dr Karimi’s work here.