How Beyond Limits Is Expanding the Limits of Cognitive AI and Product Development

Written by Ghazala Irshad
Published on Oct. 30, 2020
How Beyond Limits Is Expanding the Limits of Cognitive AI and Product Development
Brand Studio Logo
beyond limits
beyond limits

The way that Stephanie Datu oversees technical project management and innovation lives up to her employer’s name. Datu works for Beyond Limits, an LA-based startup borne out of Caltech’s Jet Propulsion Laboratory for NASA’s space program, that creates unique, industrial-grade cognitive AI for industries like energy, healthcare and fintech.

 As a technical project manager, Datu stretches in multiple directions throughout the software development lifecycle (SDLC), drawing upon her previous experience as a software engineer to ensure an agile process from proof-of-concept to product, while also pitching in to fix small bugs. 

“Beyond Limits still has that fun startup component where you have the opportunity to wear multiple hats where needed,” Datu recently told Built In LA.

As a bonus, every project stakeholder gets to be up close to the action and make an impact with their contribution, like an IP or source code. Datu said just being exposed to the company’s unique cognitive AI can inspire innovation. 

“Everyone feels like they are truly on a path of progress with cutting-edge technology at their fingertips,” Datu said.

Meanwhile, her past as a technical recruiter talking to people from diverse backgrounds comes in handy for aligning Beyond Limits’ cross-functional teams — composed of software engineers, data scientists, and AI engineers who come from a “crazy cool hodgepodge of professional experiences.”  

In fact, nobody on her team has a background in the industry that they’re currently creating technology for: energy. But instead of hampering product development, Datu said their diverse outlooks actually drive innovation, which is uniquely beneficial for the kind of complex, cognitive applications that Beyond Limits intends for its AI.

“A healthy mixture of backgrounds is an ideal scenario for creating technology like artificial intelligence,” Datu said. “You want that variety of perspectives from different specialties, opinions, and walks of life creating the technology that’s meant to make decisions in a similar manner to the human brain.”

Datu’s role is that of connective tissue — bridging gaps between software engineers, data scientists and AI engineers to fuel innovation. She holds a lot of meetings so she can always have a finger on the pulse of the project and guide seamless collaboration between teams, using tools she would “lose [her] mind without” like Atlassian Suite, JIRA, Confluence and Bitbucket. She’s also regularly fielding feedback to improve the efficiency of the process. 

Datu said a great example of successful, innovative technical project management in practice is Beyond Limits’ refinery project. 

“Every team worked together to make a whole new IP,” she said. “Our software engineers and data scientists collaborated closely to make use of this really cool technology from NASA’s Jet Propulsion Laboratory and added it to the AI we already had to solve complex problems unlike ever before.” 

On top of all this, Datu still aims to do even more, making an effort to familiarize herself with data science by sitting with the data scientists so she can learn from them and incorporate their processes into the SDLC. 

That certainly wouldn’t be possible without the healthy culture that Datu and her team have built around transparency, positivity and friendship. “Friendship is the key to my personal management style at Beyond Limits,” she said.

 

Stephanie Datu
Technical Project Manager • Beyond Limits

How did a former software engineer like yourself develop the skills to become a successful technical project manager?

I came from a startup that desperately needed a technical project manager who could also get in the trenches and contribute to the ongoing engineering work. I was offered the flexibility to trial-run a technical project manager role with the openness to return to solely software engineering if I wasn’t into it, but it felt natural and I loved it. It made sense to have the ability to do both; it actually made everything easier in the long run.
 
My transition took place over time. In the beginning, as a technical project manager, I was still doing a lot of engineering at the same time. Software engineers have the advantage of being familiar with agile methodology as part of their workflow. An engineer doesn’t have visibility into everything that a technical project manager sees, but you do have a decent understanding of processes. 
 
The fact that I was also a technical recruiter before becoming a software engineer was also a contributing factor. I was pretty acclimated to talking to a ton of different people with diverse technological backgrounds. 

 

 

What various professional backgrounds do your Beyond Limits teammates come from, and how does it help them be innovative?

It’s honestly a pretty crazy cool hodgepodge of professional experiences where none of us actually come from a background that totally lines up with the current project’s industry. The closest background that parallels our main refinery project would be our chemical engineer.

A huge benefit is that all of our software engineers working on this super advanced AI are very smart individuals. It also helps that our team has a rock star UX professional.

In all honesty, a healthy mixture of backgrounds creates an ideal scenario when it comes to creating technology like artificial intelligence. You want that variety of perspectives from various specialties, opinions, and walks of life creating the technology that’s meant to make decisions in a similar manner to the human brain.

 

What skills and knowledge can a technical project manager offer that a project manager cant?

When you’re managing a technical project — especially one that heavily involves software — you have a clear advantage if you’ve actually built these things yourself in the past, having experience with the programming that your team is dealing with every day. It’s so much easier to problem-solve with your software development team when you have that level of empathy and deep understanding of the process. I know when an engineer is truly “stuck,” they may just need an extra level of support on a difficult problem. 

It’s also an advantage to be able to actively help out when a knowledge gap may exist. It’s a good thing I’ve never fully taken off my software engineering hat. Sometimes I can contribute to helping fix small bugs, which is really fun. Beyond Limits has that fun startup environment where you have the opportunity to wear multiple hats. It’s also great when you stumble on a cross-functional opportunity for someone on your team where they could potentially thrive and contribute extraordinary value.

 

Which principles of the SDLC and project management do you believe play the most important role in transitioning a proof-of-concept to a scalable product?

When it comes to the principles of the SDLC, from the beginning, Beyond Limits has sort of mingled everything together. You need all of it and often all at once — which can blur the lines for a lot of processes — and that’s OK. Sometimes you need to carry out planning and analysis at the same time, or design and implementation at the same time. Beyond Limits understands the need for flexibility when it comes to laying out a project, but we balance that by following an agile project management environment and placing a lot of importance on timelines. We are very transparent and collaborative with our clients when it comes to project timelines.

Flexibility also comes into play when all the pieces don’t fit quite perfectly in place right out of the gate. You may not have a UX designer at the beginning of a project, so front-end engineers will fill that need in the beginning. Once the project starts moving beyond proof-of-concept is when supplementary resources start to become more readily available for design. This is the moment when you can go back in and finally perfect those portions. In the beginning, the focus should be to prove viability, then move into implementation and execution.

 

What skills and tools help you to bridge the roles of data scientists, software engineers, AI engineers and other technical roles? How do you connect everyone for cross-functional collaboration in a remote world?

In my opinion, the most important are soft skills. When you have a lot of brilliant people, you want them to feel comfortable talking to each other about their opinions and ideas. This transparency also helps my understanding of what everyone on my team is doing.

It’s important for me to learn the roles I’m not familiar with to make sure I understand enough to provide helpful support. I have a lot of experience with the software engineering side, but I’m a lot newer to the data science aspect. I make an effort to sit with them and learn their processes so we can fit that into how we implement our software.

It boils down to a healthy team culture and making sure everyone works as well as possible together. Communication, friendship, teamwork, and positivity are very important for a technical project manager to perpetuate. Friendship is the key to my personal management style at Beyond Limits.

It’s also important that the team has fun together. We were already working toward becoming cloud-agnostic before going remote, so nothing has changed too much as far as work goes. But socially, it’s been a big adjustment. Events like birthdays are a big deal to me, so I’ve tried to keep that going with secret meetings that are meant to catch the team member off guard with a surprise “Happy Birthday.”

Beyond Limits’ Favorite Cross-Functional Collaboration Tools 

  • Zoom
  • Slack
  • Confluence
  • Atlassian Suite
  • JIRA, Bitbucket
  • AWS Cloud
  • Azure

As you move through the process and more roles and teams become part of the mix, how do you maintain agile development?

Atlassian Suite has been a lifesaver; I would lose my mind without JIRA, Confluence, and Bitbucket. For me, those are the most helpful tools. On top of that, meetings — a lot of meetings — to ensure everyone is on track and no one is blocked. What’s nice about being agile is the cadence. We get into the routine of two-week sprints while also maintaining flexibility, depending on the needs of each unique project.

It’s also important to continue investing in your tech stack so that you’re always innovating and investigating, making sure it’s the best option for the project. Our software engineers are using a variety of different stacks, depending on the project: React, Java, Python and Scala. Sometimes the best solution needs to be examined and explored on a project-by-project basis. We foster a lot of freedom to pick and choose solutions and make thoughtful proposals to management depending on the project.

Even new grads may get the chance to have a closer look — and contribute to — an IP and source code.”

How do you keep cross-functional teams in sync throughout the entirety of the process from project to product?

A lot of the same tactics carry over from trying to maintain an agile development process: lots of meetings, standups, daily scrums. An abundance of communication is especially important in this virtual working world. All teams are always talking and I try to keep a finger on the pulse of all teams involved with a project. We’re always getting collaborative feedback on the processes that work and don’t work for all teams, cross-functionally. Perhaps a particular meeting series has become redundant and thus less productive, so we get rid of it. Consistently exploring the way a dynamic performs is essential to efficient processes across the entire organization.

 

Whats an example of how cross-functional work has driven innovation at Beyond Limits?

A great example of this would be Beyond Limits’ refinery project. Every team worked together to make a whole new IP. Our software engineers and data scientists collaborated closely to make use of this really cool technology from NASA’s Jet Propulsion Laboratory and added it to the AI we already had to solve complex problems unlike ever before.

Something great about Beyond Limits is that project stakeholders are generally closer to the action than they would probably be at a bigger company. Even new grads may get the chance to have a closer look — and contribute to — an IP and source code. You’re also exposed to some really cool artificial intelligence; Beyond Limits’ AI is so unique because it is symbolic and cognitive, as opposed to the more commonly developed data-driven AI techniques. This kind of insight can inspire ambitions toward a more innovative mindset because everyone feels like they are truly on a path of progress with cutting-edge technology at their fingertips.

Responses have been edited for length and clarity. Images via company.

Hiring Now
Celonis
Artificial Intelligence • Machine Learning • Software • Analytics • Business Intelligence