How three IT experts built AWS skills to tap into the power of generative AI
At Baker Tilly I am encouraged to learn and pursue AWS Certifications… That’s where I learned about all the different AWS services, as well as the pillars of the AWS Well-Architected Framework,” he says… Like many who never received formal cloud training in university programs, she did much…
Artificial intelligence (AI) itself may not be new, but the field was thrown into a new level of overdrive in late 2022 when the launch of ChatGPT brought generative AI and large language models (LLMs) quickly and irrevocably into the public’s consciousness. Over the past year, the world’s largest companies marshaled a stunning number of resources to capitalize on the AI revolution, creating massive demand for experienced AI developers. Colleges and universities are working to put together comprehensive curricula around AI, but it will take time for the graduates with AI skills to begin entering the job market.
In the meantime, IT workers across a range of disciplines are proactively upskilling to master the skills needed to harness these powerful tools, immersing themselves in hands-on experimentation. In September 2023, AWS released Amazon Bedrock, a fully managed AI service that allows AWS developers to easily integrate generative AI models into their cloud projects. We spoke with three IT professionals, participants of the AWS Community Builders program, who come from diverse places, backgrounds, and experiences. Despite their differences, all of them were able to quickly come up to speed on Amazon Bedrock and put these technologies to use, thanks to AWS’s training resources.
Neeraj Iyer: Chatbots that build chatbots
Neeraj Iyer is a senior cloud data engineer at Baker Tilly, a global tax, audit, and consulting firm with clients across a broad spectrum of industries and an AWS Partner. Though Iyer has an undergraduate degree in computer science and a master’s degree in IT and data management, during his entire formal education, he only had a single course in cloud computing.
“I heard about AWS while I was doing my master’s, but it wasn’t until joining Baker Tilly in 2019 that I finally got the chance to delve into AWS. At Baker Tilly I am encouraged to learn and pursue AWS Certifications. That’s where I learned about all the different AWS services, as well as the pillars of the AWS Well-Architected Framework,” he says.
Neeraj Iyer at AWS re:Invent 2023
Ultimately, Iyer would go on to earn three AWS Certifications, including AWS Certified Cloud Practitioner, AWS Certified Solutions Architect – Associate and AWS Certified Data Analytics – Specialty. Three years ago, he started working on chatbots for several of Baker Tilly’s clients.
“At that time, we didn’t have access to AI, so all it could do was surface basic queries from a knowledge repository,” he says. “It was very limited. If the bot couldn’t find an answer it would prompt the user to email their question, which would be replied to within a few days.
“Amazon Bedrock makes it incredibly easy to build generative AI applications. It gives you access to several different AI models from leading companies, which makes it incredibly powerful. Now I can program chatbots using a chatbot. I tell a bot what kind of bot I want to build, and I have a prototype in a few seconds. Six months ago, doing the same thing would have taken me three days. It’s a revolution in productivity.”
For his projects, Iyer also leveraged the latest Amazon QuickSight Q innovation, incorporating new LLM capabilities available through Amazon Bedrock and enabling generative Business Intelligence (BI) capabilities in QuickSight.
“At Baker Tilly, we are excited to employ generative AI in our data engineering and BI use cases with Amazon QuickSight, helping our customers gain valuable insights and make predictions from data through natural language conversations.”
Iyer credits his AWS Certifications for giving him the experience to dive into Amazon Bedrock so quickly. “My knowledge of other parts of AWS meant I could easily jump into the AWS documentation,” he says. “I’m also an AWS Community Builder, which comes with access to many other resources, including webinars, blogs by fellow community builders, and educational blogs by AWS employees.”
Amelia Hough-Ross: Transforming data insights for research
Amelia Hough-Ross is a senior cloud engineer at the Pacific Northwest National Laboratory (PNNL), a branch of the U.S. Department of Energy. With more than 5,000 researchers working on a dizzying array of topics, from nuclear waste cleanup to computational chemistry, the lab is an extremely data-intensive place.
Amelia Hough-Ross at AWS re:Invent 2023
After getting her undergraduate degree in computer science, Hough-Ross worked in conventional data centers before landing a job with a cloud-focused company in 2015. Like many who never received formal cloud training in university programs, she did much of her learning on the job, augmented by whatever other knowledge she could pick up online. She scheduled and took her first AWS Certified Solutions Architect – Associate exam during re:Invent 2015, only to find herself bewildered by the questions.
“It was an epic fail,” she says. “And that was great. It told me all the things I didn’t know.”
After consulting with some of her colleagues, Hough-Ross put a more structured training plan in place for herself, and ultimately passed all three Associate-level AWS Certifications, along with the AWS Certified Security – Specialty exam. In 2021, she was studying for the AWS Certified Solutions Architect – Professional exam when AWS Skill Builder, AWS’s online learning center with 600+ free digital training courses and AWS Certification exam preparation resources, was released.
“AWS Skill Builder was very helpful for validating that I was on the right track,” she says. “The internet is full of stuff, but it’s not always the right stuff.”
Today, Hough-Ross helps the scientists at PNNL figure out how they can use cloud resources safely and effectively in their work.
“We have several hundred people using the cloud in various ways, and they’ve generated thousands of support tickets. If we could query that data set for insights, it would allow us to see trends in what our customers are asking for and help us get ahead of their needs. I thought AI natural language processing might be able to give me a synopsis, but all the text-parsing products I tried produced substandard results and I didn’t have the time to train my own AI model just for this,” she says.
Cue Amazon Bedrock.
“It had pretrained models, which was exactly what I needed,” Hough-Ross says. “One of the big advantages of Amazon Bedrock is that it incorporates several different pretrained models from the leading companies in AI research. “It lets you pick and choose which one works best for your use case.”
Hough-Ross’s proficiency allowed her to dive in and get going with Amazon Bedrock very quickly after taking an overview course on Bedrock. “Earning AWS Certifications and easy access to AWS learning resources prepared me to understand cloud intricacies. I can confidently navigate and use advanced products like Amazon Bedrock supporting my lab’s researchers in leveraging emerging technologies.”
Renaldi Gondosubroto: Virtual streamers and AI therapy
When Renaldi Gondosubroto received his master’s degree in computer science, he also knew a fair amount about the cloud. But not because of anything he had learned in school. Instead, he learned how to work in AWS without formal training by doing freelance IT work during college for clients that used AWS. He didn’t have time to study for AWS Certifications on top of his classwork and jobs, but once he graduated in 2020, he wasted no time. By the end of 2021, he had passed every certification exam offered by AWS.
Gondosubroto began his AWS learning using a combination of books and online resources. AWS Skill Builder was introduced while he was still on his learning journey in 2021, and he used it as a resource for several of his later exams. Now a software engineer at SEEK, an Australian employment search company, he spends his spare time working on ambitious projects that push the boundaries of what AI can do.
Gondosubroto credits his prior AWS training with being able to ramp up so quickly on Amazon Bedrock. “When you start to build more scalable applications, you really need to have a core foundation of AWS knowledge.” He says the AWS Solutions Architect – Associate certification was particularly helpful.
“It’s great for building a fundamental knowledge of how AWS works. Training for AWS Certifications also gives you the patience to dive into documentation for new features. You have to have a continuous learning mindset, listen for new innovations, learn from the community, and go to events.”
Renaldi Gondosubroto at AWS re:Invent 2023
At AWS re:Invent 2023, Gondosubroto presented one of his projects, an interactive AI-based Twitch streamer. It’s a bot combined with an animated avatar that can reply aloud, complete with lip-synch, in response to typed input from users.
“It all came from my interest in experimenting with event-driven architectures in Amazon Bedrock,” he says. “How can the bot respond to different events? How can it respond differently to different people? Could I incorporate all of this into a platform such as Twitch? The result was a completely autonomous, synthetic Twitch streamer.” One of the bot’s specialties is playing chess with visitors. When in tutorial mode, it will even offer constructive criticism on your last move.
Gondosubroto also created an app intended to help people maintain good physical and mental health over the holidays. It includes a virtual therapist, a nutrition information app, and a virtual journal with sentiment analysis to gauge the user’s mood. The therapist was built using Claude, one of the pre-trained AI models available in Amazon Bedrock.
Gondosubroto sees no end to the opportunities AI will bring for those willing to educate themselves with the resources AWS offers.
“Right now, the world is in an adaptation phase. This boom is real, and there are not enough qualified people to feed the market. The universities are doing their best, but they have not caught up to this yet. There are already new job titles that have never existed before, things like Director of Generative AI or Generative AI Development lead. It’s definitely a great time to start learning how to work with generative AI.”
To learn more about getting started with Amazon Bedrock, check out our free courses in AWS Skill Builder:
Author: Training and Certification Blog Editor