Results
"AI shouldn’t be seen as the ultimate solution. It should be a tool that helps you get to the solution."
Arun P.
Data Scientist and AI Engineer

What’s the ideal career for someone with the soul of an artist and the mind of a mathematician? For London-based Arun Godwin Patel, it’s being a freelance generative AI specialist. Generative AI is a type of AI that creates new content based on a variety of input (e.g., text, images, sound) by learning patterns and structure of the data it’s trained on. And Arun is so skilled at it that companies globally invite him to work on their most innovative projects.

Arun started in the field when AI was so new, it was seen as the stuff of science fiction. He was a university math student at the time interning at SAP, a multinational software company. “The internship gave me insight into how math could be used in real-world business applications, and I was intrigued.” He was so hooked that he got a master's degree in AI from the University of Surrey.

Since then, he’s been using AI, data science and machine learning (ML) to help companies solve real-world business problems. In 2020, he left the corporate world to offer his expertise as an independent professional on Upwork. We hopped on a video call with Arun to chat about how businesses are using AI, misunderstandings about the tech, and what it’s like to freelance in the field.

Arun

How companies are using AI

The type of project inquiries he gets “depends on what’s hot at the moment.” At the time of this interview, that’s been developing recommendation systems and conducting text data mining.

His clients range from advanced AI users who built their products and services around AI, to those who want to see what’s possible with the tech but usually aren’t ready to launch a product yet.

This was the case for one recent client who sought to test whether it was technically feasible to create a system combining virtual reality (VR), AI, and sensors to provide a personalized mental health management system.

For this project, Arun was tasked with building a dynamic VR environment that identified the user’s “mental state” by reading their heart rate variability (HRV). Based on their data, the user was placed into a soothing virtual experience that was optimized for them within seconds. As the user’s HRV changed, the environment subtly adjusted elements including volume, sunlight, and speed of movement to provide the most beneficial experience.

VR

Arun is quick to add that not all companies know what they want to do with AI. Many want to jump on the AI bandwagon without a clear understanding of how to use the technology effectively. In these cases, Arun generously takes time to guide them in setting realistic expectations and defining what a successful outcome looks like. “I want to be helpful. Sometimes, that means suggesting they don’t pursue AI yet,” he said. Taking this thought further, Arun believes that…

AI shouldn’t be “the” solution

“People get very excited about generative AI and their expectations can be a little bit skewed,” said Arun. If a company gives him a use case for AI, but from experience he knows that it’s not a good use of the technology, he’ll tell them. “I want to do right by them. I’d rather turn away an opportunity than build something that is not the right solution for them.”

He explains, “Using AI is kind of like swinging a sledgehammer on something that may only need a very light tap. AI shouldn’t be seen as the ultimate solution. It should be a tool that helps you get to the solution. If a problem can be solved with a spreadsheet or with some simple rules, there's no point in using AI.”

AI shouldn’t be seen as the ultimate solution.
It should be a tool that helps you get to the solution
.”

“You have to think, How will this algorithm influence the business and deliver us value?” said Arun. At times, companies miss the real value AI can deliver. He provided this illustration: An employee takes eight hours to complete an analytical task at 70% accuracy. Many people would focus on using AI to achieve 100% accuracy. But what if using AI could maintain a 70% accuracy rate and complete the task in five minutes? Then the employee can spend eight hours doing something that may support business growth.

In his own business, Arun judiciously uses AI tools to speed up production. “AI tools are the future, so I like to keep up to date with them,” he said. But the technology is still new enough that he remains cautious. “I prefer diving into the details rather than relying on an automated system to do it.” He doesn’t use AI to write entire blocks of code, but he uses the tech for auto-completing lines and suggestions. “Even though the time saving is minimal, it makes programming more efficient,” said Arun.

Why he likes freelancing

“The thing I enjoy most is taking a client’s vision and figuring out how to actually make it work,” said Arun. He did that in the corporate world and appreciated the experience, but the work wasn’t satisfying. “The projects were so big and I had such a small part in them that I didn’t feel like I made a tangible impact.”

The thing I enjoy most is taking a client’s vision and figuring out how to actually make it work.”

After four years, he decided to freelance full time. “I really wanted to focus on doing things where I can actually see how my work will have a positive impact on the business or—even better—humanity.”

Arun launched his freelance career by joining Upwork. “Upwork increases my exposure to a variety of companies and cutting-edge use cases. Since joining, it snowballed.” Freelancing also gives him control. “I like deciding what projects I'm going to work on. I really enjoy waking up knowing what I’m going to work on that day.”

Arun also contracts skilled talent through Upwork. In addition to freelancing, he is the founder and director of Halo Technology Lab, a research and development company that’s focused on emerging technology and human-centric use cases. He is also the co-founder and CTO of Audico, an end-to-end voice platform for hospitality, health, and home. “I haven’t used other platforms to find work, but I’ve contracted people from other platforms, which can be a bit chaotic. With Upwork, I get a good balance between a wide variety and good quality talent.”

Clients get more than an AI specialist

The field of AI is evolving so quickly that Arun believes “the best way to learn is by doing.”  As a freelancer, he has wider and deeper knowledge than many in his field because he works with several companies in various industries. The variety exposes him to “very experimental projects” that require applying creative problem solving and advanced solutions. Many times, the experience from one company provides insights that help him efficiently solve problems for another company.

His strongest advantage over other AI specialists comes from the trifecta of being a creative, CTO, and freelancer. Instead of contracting two or three people to handle different parts of a project, Arun can handle most of it. Because he can build the entire life cycle from ideation to experimentation and then production.

Advice for someone considering a career path in ML or AI

Whether freelance or full time, if you’re considering a career in ML or AI, Arun suggests the following:

  • Be persistent. There are so many concepts and technologies to learn, it’s important to start without expecting to know everything there is to know. He stresses to be patient, take your time, and invest in your skills.
  • Focus on the business value. Instead of learning specific techniques, algorithms, or models, focus on use cases because you’ll rarely work on two projects that are exactly the same. “You’ll always have to tailor something. It is more valuable to approach a problem with a general framework rather than try to fit a specific algorithm to a use case.”
  • Brush up on your hard and soft skills. The better your technical skills, the more efficiently you’ll work. But don’t overlook soft skills including communication and project management. Soft skills are just as important to a project’s success. “Most AI and ML projects fail not due to technical reasons, but due to business reasons.” The most common are:
    • A disconnect between the technical and business teams
    • The project doesn’t demonstrate enough business value
    • A lack of project direction or purpose

His dream AI project

Although Arun enjoys concentrating on projects that benefit business or humanity, his dream AI project is more lighthearted.

A self-proclaimed “massive football fan” and video gamer, Arun thinks that most football video games (soccer, to the Americans in our audience) are “outdated and rubbish.” He’d love an opportunity to build a game using AI to make the players’ movements and facial expressions more realistic when they’re on the field and not playing, such as when they’re waiting for a ball.

“If somebody came to me with a project around that, or any sport, that would be really cool… Yeah, I'd love that.”

If you have a project like that, please contact Arun immediately. But if you have another project in mind, or just want to ask a few questions, he’d welcome hearing from you about those, too.

Industry
Category
Employees
Join the world's work marketplace

Recommended success stories

Join the world's work marketplace

Find great talent. Find great work. Are you ready to move your business or career forward?

Find talent
Find work
Get Started