AI Team Spotlight

Meet the AI team here at Posh! Consisting of engineers and researchers that power our bots with machine learning to understand user's problems so that they can help solve them. 

On the team are: Rishi Sharma, Senior Director, AI, Dhairya Dalal, Senior Research Scientist, Michelle Fullwood, Principal Research Scientist, Stefaan Heyvaert, Principal AI Engineer, Cheyanne Baird, Research Scientist II, Katie Krajovic, Research Scientist II, Sam Tureski, AI Engineer II, and Kelley Lynch, Research Scientist. 

Read on to find out what they have to say about working at Posh, the future of AI, and more! 

What is the most exciting part of your job? 

Kelley: Getting to keep up with all the things going on in the field, new techniques, new products, new companies in the space. Getting to follow everything that’s changing in our field. 

‍

Katie: I don’t work on the same things every day, the problems I’m working on change all the time. It’s fun and exciting to approach new problems and learn new techniques. 

‍

Dhairya: The field of AI is rapidly evolving, being a research scientist you get to be hands-on and vetting new ways to be deployed. We’re at the forefront of research ideas and building products and services around them. Being at the cutting edge of these technologies. It’s fun to work with the team, we have a lot of different experiences and backgrounds (linguists, NLP…), everyone is just super nice. Really excited about the work we do. We are all nerdy in different ways and excited about the work we do! 

‍

Michelle: Whenever we deploy new features, we do everything from research to engineering to analysis. I really enjoy the process because it’s so holistic. 

‍

Cheyanne: I love keeping up with what’s going on in the field. We’ve all been doing great experiments with language models. I like diving into the data and analysis and I get to do that a lot with annotation and dialogue code. Looking at how things are implemented and how they’re working. 

‍

Sam: I write and work on a lot of code and our team has really high standards. I enjoy being able to hone my software design skills and I get to think hard about finding the best solution. 

‍

What have you learned in your role at Posh so far?

Kelley: Coming from academia, I’ve learned a lot about infrastructure and workflows that are necessary to keep things working in a real-world setting. In academia sometimes it’s fine if things break but it’s been a transition getting acquainted with all of our systems. 

‍

Katie: I’ve learned how to condense information down, like if I’m conducting a bunch of research I need to discern what is the most pertinent information to convey to our Product team or others and that’s something I continue to work on and get to practice a lot here! 

‍

Cheyanne: It’s been great to learn about chatbot development and conversation design, that’s been really new for me and it’s exciting to dive into that field. 

‍

Dhairya: One of the interesting things about AI products and features is understanding all the things that it takes to get a product up and running. The infrastructure, data, it’s an entire team effort that makes it possible. What’s been interesting working at Posh has been seeing the evolution of how our product has grown. We went from having a small thing to the deployment of our current offerings with complex pieces of logic. It’s a study in complexity; as you start to build more sophisticated features it requires more scalability. Understanding the role of our group in the company as a whole is fascinating! 

‍

Michelle: This is my first role where I've seen the system grow from its beginnings to something very large, and where I have to think in terms of the entire system and not just my small part of it. I'm expected to express my technical views across teams that will influence how they will work, which has been new and exciting to me.

‍

The team in our Boston office. 

‍

What’s something that makes you proud to work at Posh? 

Michelle: I really like the way our team has grown. We went from three people to the most diverse technical team in the company. Being diverse means that we aren't constrained to a single approach to anything - we will use patterns derived from data and deep learning and LLMs, whatever makes the product better. Diversity of thought is really important in a technical setting.

‍

Kelley: It’s really comfortable to ask questions of the team and express confusion, compared to other teams I’ve been on I’m much more comfortable here asking questions. The diversity of our team fosters an environment where I can speak up. 

‍

Sam: Because it feels so comfortable on our team to ask questions and talk to each other, I’ve grown personally and professionally by sharing my work and with asking questions. In the past before I worked at Posh I was scared to share my work, but I love my team here and I want to give them the best work I can. 

‍

Katie: If I’m working on something by myself or I’m stuck, I feel like I can talk to anyone on our team and they’ll have a different perspective that I’ve never thought of before. If I want more than one perspective I can talk to more than one person on the team and they’ll have even more ideas. 

‍

What do you hope to accomplish in your field as Posh grows and the industry changes? What are you looking forward to the most at Posh?

Michelle: Even though my title is Research Scientist, I’m very product-focused. I just want to build the best, most helpful chatbot possible. I don’t really care what powers it, I’m more concerned with how well it works.

‍

Katie: In my career, I just hope to get to build things that help people. Technology is developing and that’s awesome, but for me it just means more tools I can use to build things that will help people. That’s what I’m excited about in this field. 

‍

Dhairya: I’m focused on research and looking at these technologies and applications in different areas. What I find fascinating is the evolution of problem-specific solutions and scaling them out. Language models had been specifically focused on predicting intents and we rolled out our models. We had one model for intent recognition, custom intents, another for NLP approaches. One of the cool things about this space is that we’re truly moving past having these solution specific architectures and having something more universal. As you scale the solutions, you scale the types of problems you can tackle. So I’m really excited about the new opportunities for us to tackle and thinking about how AI can help with those! Especially with things like financial literacy and core banking. This means we can build more and more exciting models and provide more value to our customers. 

‍

Kelley, Michelle and Katie

‍

What would you say to someone who is skeptical of AI?

Dhairya: It’s good to have a healthy skepticism because it allows you to think about how the solutions will be used. What separates AI research at Posh from other places is that we are very deliberate in how we use these technologies in a responsible way. You won’t see our chat bots giving terrible advice or have awful responses because we use our technology in a responsible way. We don’t make haphazard decisions, we are always thinking about our customer experience first and foremost.

‍

Final thoughts and fun facts!

Sam: I love my team! 

‍

Katie: We have wug emojis in Slack. Wugs are a linguistics thing and they’re really great and we add more anytime they become relevant! We have a whole document called the Wug creation station. We have like 15 or 16 now. 

‍Dhairya: We have a really creative team and we leave easter eggs across the company Slack.