Behind the Code: Meet Laura Chiticariu

An IBMer for over a decade, Laura Chiticariu has played a significant role in developing AI technology for the company, particularly in Natural Language Processing. Now a Chief Architect of Knowledge and Language Foundation for IBM Watson and Cloud, she is diving deeper into the exploration of NLP and further build out its capabilities.

Learn more about Laura and her journey in the latest edition of our “Behind the Code” series.

Tell me a little bit about your role. What do you do as Chief Architect for Watson Knowledge and Language Foundation? How long have you been at IBM?

I joined IBM Research in 2008, after obtaining a Ph.D. in Computer Science from University of California, Santa Cruz. My expertise is in Natural Language Processing (NLP). After 9 years in Research, I decided to move to product development, first as NLP squad lead for Watson Compare & Comply, and later Chief Architect for Watson Natural Language Understanding.

I have recently taken on the role of Chief Architect, Watson Knowledge and Language Foundation, with technical leadership over Watson Natural Language Understanding, Watson Knowledge Studio and Watson Knowledge Graph. In this role, my responsibilities include standardizing and expanding the language coverage of Watson’s base NLP layer, including primitives such as tokenization, part of speech tagging, dependency parsing, shallow semantic parsing and core algorithms for various text extraction and classification tasks, and driving adoption of these base NLP capabilities across IBM Data & AI offerings.

What led you to pursue programming as a career?

My first encounter with computers was in primary school – my mother was leading the Center of Informatics in a textile factory, and she is my inspiration. She would sometimes bring both me and my sister along, and we’d play Space Invaders on Z80 computers. I took my first programming class in 9th grade, which coincided with my parents purchasing our first PC, an i386.

At first, I found the subject difficult. However, I had an excellent professor; her style of teaching and dedication enabled me to overcome my initial difficulties and simply put, I became enamored with programming. And just like that, I was one of those fortunate enough to have figured out exactly what I wanted to pursue as a career very early on.

What is an exciting project or initiative that you had the chance to work on at IBM?

SystemT, the world’s first declarative system for expressing, optimizing and executing NLP algorithms. SystemT was born in IBM Research – Almaden in mid-2000s, and took a completely different approach to text analysis compared to all earlier systems. Similar to core principles in relational database systems, SystemT enables the developer to express “what” to extract from unstructured text, in a declarative fashion, and lets the optimizer decide “how” to execute the specification. Today, SystemT powers various capabilities in multiple IBM products, including Watson NLU and Watson Knowledge Studio, two of the offerings under my current technical leadership.

I have been a core member of the SystemT research team, and later the development teams, throughout my entire career at IBM, contributing to every aspect of the system, resulting in both internal impact: runtime system building, hardware optimization, developer tooling, working with clients, and transfer to IBM product groups, as well as external impact: research publications and tutorials in top conferences, and teaching graduate-level classes in universities. Building SystemT really challenged me to understand NLP at the lowest level of detail.

What were some of the challenges you’ve faced in your line of work? Have you ever faced any particular obstacle because of your background or gender?

The challenges I face in my job are primarily of technical nature. Solving problems is a natural part of any engineering discipline, and as they say, if it were easy, someone else would have solved it already.

I have been fortunate enough not to have personally encountered challenges due to my background or gender. All my managers and mentors in IBM have been extremely supportive, and in fact, IBM has a great track record of mentoring and promoting a diverse workforce.

That said, women and minorities are underrepresented in Computer Science and related fields. This is a fact. Therefore, it is important for each of us to do our part in improving this status quo, and contribute, in any way we can, to increasing diversity at all stages in the pipeline, from primary school, to higher education, to professional careers. On my part, throughout graduate school and during my career in IBM, I have been teaching programming workshops to middle schoolers (both girls and boys) in disadvantaged school districts.

In your field, is there a common misperception that you’d like to correct?

AI is not magic.

What advice do you have for young women who want to learn to code?

My advice would sum up as: get an introduction to the field and seek like-minded individuals. A good resource to get started with programming, as well as finding women role models and mentors, is your local SWE (Society of Women Engineers) chapter. SWE chapters are present in a majority of US universities. They organize local programming workshops, as well as various mentorship and outreach activities. This will give you a chance to understand the various flavors of computer programming (e.g., web developer, game developer, UX/UI developer, etc).

After that, a natural next step would be attending online classes and local meet-ups related to your areas of interest. This will give you a chance to meet others getting started in the field, and form relationships.

Finally, this is advice that applies to starting in any field, not just programming: be curious, work hard, and don’t get discouraged.