Coding Bootcamp Instructor User Stories: Enhanced Learning with AI

Jonathon Hinchley is a full-stack software engineer and coding bootcamp instructor with a decade of experience. In addition to teaching, he specializes in website development and APIs and currently uses AI assistance for writing unit tests and asking clarifying questions about code. Pieces for Developers, especially Pieces Copilot, impressed Jonathon and earned a consistent place in his workflow. Let’s dive into why.

Benefits Highlights:

‍

Streamlined Workflow with Small Efficiencies: Jonathon Hinchley, a full-stack software engineer, integrates Pieces into his workflow to simplify tasks like snippet storage and code reuse. By centralizing snippets and providing accessibility across platforms, Pieces significantly reduces the time and frustration associated with locating and utilizing code snippets, resulting in a more efficient development process.
Enhanced Learning and Collaboration for Students: Pieces offers features like code-specific OCR and Copilot Chats, which not only assist developers in their coding tasks but also facilitates learning and collaboration among students. By enabling students to ask questions, seek explanations, and receive feedback on code, Pieces promotes critical thinking and skill advancement in a supportive learning environment.
Versatility and Security in Offline Usage: Pieces Copilot's offline functionality is highlighted as a game-changer by Jonathon, allowing developers to remain productive even without internet access. This feature not only enhances mobility but also ensures data security, making it particularly valuable for developers working in security-focused environments where offline capabilities are essential. Additionally, the ability to leverage Pieces for local searches and understanding legacy code further demonstrates its versatility and potential impact across various development scenarios.

Jonathon’s Tech Stack:

Languages: .NET, C#, SQL, JavaScript, HTML, CSS
IDEs: VS Code, VS Enterprise
Other Tools: Pieces, MongoDB

Small Efficiencies Make a Big Difference

Pieces initially earned its place in Jonathon’s stack the same way it earned its place in thousands of other devs’ stacks: by simplifying the mind-bendingly complex task of building software. “Just the ability to save a snippet was huge,” Jonathon explained, “because I write a lot of code where I can reuse large chunks of it. Before, I would dump my code into a single-page repo in GitHub. Whenever I needed it, I would grab it from that repo, but it's a little frustration that becomes large when you're trying to code things quickly and deliver fast.” By storing all of his snippets in one place that’s accessible from his browser, desktop, and IDE, Pieces vastly simplified Jonathon’s snippet-reuse flow.

When asked about how he uses AI in his coding routine, Jonathon says, “I'll simplify the 20-minute workflow into two minutes, which is what Pieces did.” From adding (and re-adding) context to ChatGPT or GitHub Copilot, Jonathon went to using Pieces as just one tool for all of his AI help. “I would switch between Visual Studio and Visual Studio Code, but GitHub Copilot doesn't work that well in Visual Studio. Pieces is nice because it's one tab. It just stays on my left monitor that I'm looking at. It's never closed anymore, it just stays open quietly. I think it's under 4 MB of memory. I don't even think about it, because I have 20 GB— Pieces just runs in memory.”

Using Pieces between editors was a huge time-saver for Jonathon, and made for a big reduction in context-switching. “I'm switching editors so much; I can't be everywhere at once,” he says. “One of the things I used Pieces for was Docker. I was writing a Docker file to build this API as a Docker container and I was port mapping, where you have a host machine and a client machine. I had forgotten which side was the host of the client. I wondered if Pieces would know, so I just typed it and it was there!” Jonathon was blown away by the help Pieces provides. “It's like little efficiencies like that. When you don't know all of the things you need, it's like you have a one-stop shop.”

Don’t Retype Code, Use Pieces OCR

One of Jonathon’s favorite Pieces features is truly an unsung hero— code-specific OCR. “I think it’s life-changing to combine two AIs and gain the ability to understand my code and also the OCR from code images,” Jonathon says. Pieces’ OCR extracts the code from an image or screenshot and allows you to edit or reuse that code without retyping it.

In Pieces Copilot, you can even drag an image into a copilot chat and start asking questions about the code right away. “To replicate that interaction, I would have had to paste three times into ChatGPT,” says Jonathon. “I would have to extract the code myself from a PNG by manually typing it, probably make a typo somewhere, put it in ChatGPT, and get a confusing answer.” Pieces OCR and Pieces Copilot dropped that three-step interaction into one.

Pieces OCR isn’t just useful in copilot chats for individual developers. Jonathon also sees huge benefits for his boot camp students. “If I'm presenting on PowerPoint, I don't even have to copy code to them. If I pause for a second, they can screenshot my screen with the code snippet, throw it into their Pieces, and I don't have to copy and paste my code over,” he explains. “Pieces explains the code and talks them through every line. I'm not going to do it. I'll let Pieces do that.”

Pieces is Great for Students, Too

OCR is not the only way Pieces can help students learn more efficiently. While generative AI is making huge strides in efficiency and reliability, it’s certainly not perfect. “I have definitely seen students use ChatGPT and they create things that hallucinate,” Jonathon says. He does not think that students learning to code should trust AI implicitly; he’s advocating for critical thinking and problem-solving. “I'm trying to teach them that if you're going to use AI, don't use it to write the code,” he explains. “Use it like I am—maybe it's unit tests, maybe it's mock data, maybe it's a conversation.”

Jonathon gives a real-life example of how a student can effectively leverage Pieces. “Imagine we're reviewing a pull request to merge their code into the repository,” he says. “We're saying, ‘Hey, what’s wrong with your code?’ With Pieces, it's like having a Jonathon there saying, ‘Let's look at your code together. Let's have an objective eye towards what this code is doing.’” Students can ask Pieces to explain code or give them feedback on something they’ve built instead of waiting for an in-person instructor to get to them. Learning to ask and answer these questions ensures that students are thinking critically and advancing their skills.

If you’re a student yourself, check out these resources on how you can use Pieces in your workflow:

Get The Answers You Need with Pieces Copilot

Personally, Jonathon’s favorite Pieces feature is Pieces Copilot. “I save useful snippets, but to be honest with you, what I end up doing more than saving snippets is the Pieces Copilot chats.” Pieces Copilot is useful, yes, but it’s also a great coworker for a solo dev. “This is my favorite part of AI, is the little conversations that I have. They're fun to talk to! Sometimes it has some personality to it,” he explains.

“Not a lot is going right today, but thank you Pieces for being so kind—your grammar is nice. You're not trying to be mean, you're not mad at me. It's great that it's emotionless. It's just giving you exactly what you asked for. That, to me, is part of it; it's an assistant or companion that doesn't have emotion. It doesn't get angry. It doesn't get tired. If you're like me and you're coding at 2 a.m. trying to get work done for a client, thinking, ‘I need this answer, I wonder if Pieces has it.’ More often than not, it has it.”

Pieces Copilot is useful for more than a friendly response in the middle of the night. Jonathon also appreciates that Pieces handles context well and often uses copilot in the browser to learn more about a site without reading it himself. “If I'm on Microsoft, I'm trying to do some app service things on Azure, and I can’t find the answer to a question such as how does billing work if I scale down the app service? When does my compute hour end? How is that calculated so I know when my billing is done? Because I'm spending thousands of dollars a month. How do I know where the line is drawn? Rather than engage with Microsoft support, I’ll ask Pieces,” he says.

Another game-changing feature of Pieces Copilot? It uses offline AI. For Jonathon, this is huge. Practically, it’s great for travel, he explains. “If I'm going out to Florida, I can jump on the plane, keep coding, and I don't have to use the cloud for my AI. I can turn on airplane mode and still get the feature done, land, turn on my mobile hotspot, and deliver the code,” he says. “To me, that’s one of the beautiful things that something like GitHub Copilot doesn't have yet, is the ability to go offline.” See why thousands of developers, including Jonathon, use Pieces as a GitHub Copilot alternative.

Using Pieces Copilot locally and offline isn’t just great for mobility, it’s also secure. “If it's on my computer, I unplug the ethernet cord. No one is getting your data; you installed Pieces,” he says. “No one else has my data now. I'm air-gapped from the internet, no one can touch my stuff.”

While he doesn’t currently work in an extremely security-focused field, Jonathon sees the potential for devs who do. “This application helps these developers to be more productive. In a legacy environment like the IRS or a government contractor, you're going to have way more legacy code that the people have retired from, right?

I've seen developers who worked there for 20 years retire. Now the young devs that just graduated from college have to understand 20 years' worth of Cobalt code that they never learned in college,” he says. Luckily, those developers could ask Pieces. “Hey Pieces, I'm not allowed to go to the internet, just do a local search. Here's this Cobalt code. Can you understand it? Okay. Let me ask you this question about this 3,000-line method that I don't understand. Let's break it down into smaller pieces.” The possibilities are endless.

Final Thoughts

Jonathon knows that AI assistance is the future of software development, even if it has its drawbacks. Pieces is a favorite AI dev tool because it doesn’t feel like a robot— chatting with Pieces Copilot feels natural. “It feels like I'm talking to another developer. It’s like the scene in The Wizard of Oz— I pull back the curtain and there’s a developer in there,” he says. “I could say more, but I think that the best summary of Pieces is it feels like another developer on the other side of the screen.”

‍