Share:
Share:
Breaking into machine learning often starts with curiosity, but it is sustained by continuous learning and hands-on experience. In this Coffee Chat, we speak with Jay Jha, a Computer Science Fellow whose work spans natural language processing, generative AI, and real-world applications across healthcare and finance.
From analyzing social media data in graduate school to building large language models for radiology and financial text, Jay’s career reflects a consistent theme: learning by taking on complex, real-world problems. He stays current in a fast-moving field not by trying to follow everything, but by building, experimenting, and working directly with new tools and technologies.
In this conversation, Jay shares how students can move from theory to practice, the habits that help them stay adaptable, and why developing confidence in machine learning comes from learning to navigate ambiguity rather than avoiding it.
A: My journey into machine learning began in graduate school, where I saw how powerful NLP (natural language processing) can be for understanding real-world behavior. During my graduate research, I worked on analyzing social media data to understand addiction patterns and public health trends. That experience demonstrated the real world impact of NLP, and I knew I wanted to continue working in this field.
From there, my career kept moving toward harder and more applied challenges. At Rad AI, I helped build a large generative model that writes radiology impressions, and at Alexandria Technology I have worked on financial NLP, large language models, and speech recognition systems. Each step was a chance to work on real data, real products, and problems where the impact is clear.
My path has never been a straight line. I learned ML by experimenting, breaking things, and building projects that felt slightly beyond my skill level at the time. That habit of learning through building is what brought me into the field and what keeps me excited today.
A: I stay up to date by working on real projects that force me to learn. Most of my growth comes from building systems and solving problems I have never solved before, whether it is training large NLP models, evaluating new architectures, or deploying pipelines.
I also follow a few trusted sources instead of trying to keep up with everything. I read top ML conference papers, check release notes from major model and framework updates, and follow engineers and researchers on social media who share practical insights and experiments. Those short, real-time updates often highlight trends long before formal publications catch up.
The most important habit is staying hands-on. I regularly test new models, try new tools, and run small experiments to see how things behave in practice. When you learn by building, you never feel behind because each project teaches you something new.
A: The biggest shift students will need to make transitioning from the classroom to the workplace is to move from following instructions to owning the problem. In my radiology NLP project, students set up the tools and software they need to run their projects , explore messy data, train models, build pipelines, and deploy real tools. Nothing is handed to them in a perfect form, and that is what makes the experience valuable.
The students who grow the fastest develop three things:
Once students start thinking in this practical, problem-driven way, they transition naturally from theory to real engineering work.
A: Moving from academic exercises to real-world projects can be challenging because real-world work rarely follows clean instructions. Data may be inconsistent, requirements may change, and systems may not work as expected. The key is learning to stay calm and approach every peoblem methodically using structured problem-solving approach to identify the root cause.
In one of my Projects, a student spent several days trying to run a simple classifier because a required software library kept installing the wrong version. Once the student learned to troubleshoot step by step by isolating the issue, reading error messages carefully, and fixing the library version, the problem was resolved. That process matters more than any single tool.
My practical advice:
Confidence comes from knowing you can figure things out, not from things working perfectly the first time.
A: The biggest lesson I wish I understood earlier is that your career grows the fastest when you are willing to learn things you are not good at yet. Every major step in my career came from taking on something slightly uncomfortable, whether it was helping build a LLM for radiology, fine-tuning generative models for financial text, or leading full ML pipelines instead of staying focused only on modeling.
Early on, I thought I had to master everything before trying it. Progress comes from starting before you feel ready, breaking problems into small pieces, and learning what you need along the way. You build adaptability by shipping things, getting feedback, and constantly refining your approach.
Continuous learning is not about collecting certificates or perfectly planned paths. It is about staying curious, taking ownership when you do not know something, and viewing every challenge as a chance to expand your comfort zone. If you build that mindset, everything else becomes much easier.
Final Thoughts
Jay’s insights reinforce a powerful idea: confidence in computer science comes from learning to work through uncertainty, not avoiding it. By embracing ambiguity, debugging systematically, and thinking end to end (from data to deployment) students develop the habits that make them industry-ready.
Whether it’s experimenting with new models, tackling unfamiliar problems, or learning through hands-on projects, Jay emphasizes that growth happens by starting before you feel fully prepared. For students preparing for careers in machine learning and AI, this Coffee Chat is a reminder that the most valuable skill of all is the ability to learn by building — and to keep building as the field evolves.
Danila Blanco Travnicek is the Director of Program Strategy & Evaluation at The Build Fellowship where she leads the education programming and its initiatives. She is a social entrepreneur who has been working tirelessly for over 10 years in the nonprofit sector to ensure more people have access to quality education. Danila holds a B.A in Business Management and a master's degree in Teaching and Nonprofit Management. She is a Professor at the University of Buenos Aires, an international speaker and facilitator and has managed and led programs with social impact in Latin America, U.S., Europe and Asia. She also received scholarships to study abroad in Finland, China and the United States.
Rik Abels is a Finance Build Fellow at Open Avenues Foundation, where he works with students leading projects in investment analysis and entrepreneurship. Rik is a Principal at Clew Partners, where he focuses on advising buy-side clients on M&A transactions, sourcing and facilitating acquisitions for strategic investors across a number of different industries.
Rik has over three years of experience in the finance field. With a background in technology venture capital, government, and public policy, he has spent the past years advising clients on acquisitions in sectors such as IT services, pharma services, and residential and commercial home services. Rik holds a Bachelor of Arts in Economics and Government from Dartmouth College.
Stay updated on all things Open Avenues!
