Trevor Taylor

💻 Are We Overhyping 'Agents'? Let’s Break It Down 🐾 User agents—like your trusty web browser—have been quietly doing the heavy lifting for years, fetching cat pictures and making life easier. So why is "agent" suddenly a buzzword? It’s the tech industry's favorite game: take something old, slap on shiny LLM-powered features, and call it revolutionary. In my latest article, I unpack what an agent really is (spoiler: it's software acting on behalf of a user) and why LLMs have people buzzing. By using semantic reasoning and tools, these smarter agents bring exciting possibilities, like making “Find cute black kittens” requests less frustrating—and occasionally hilarious when they hallucinate non-existent cats. While there’s potential here, we need to cut through the jargon and align on what agents can truly achieve. Dive into my breakdown of agents, LLMs, and why sometimes the best use of tech is still looking up adorable cat pictures. 🐱 👉 [Read the full article and join the conversation!] https://lnkd.in/gmH55Zki #AI #TechTrends #SoftwareDevelopment #LLM

3

KLAS Recently released a study collecting survey data from 500,000 clinicians regarding EHR satisfaction. Here are the top 6 learnings for EHR satisfaction. 1. Continued relevance of the 3 pillars. The three pillars are that end users need to be trained, they need to have a supportive organization, and they need to be able to personalize their system. 2. Educate beyond the basics. Educate more than what each user "button" does. It’s important to show clinicians how the EHR can support them in their efforts to better provide care. Additionally, it is important to make the education you share as specialty specific as possible. 3. Clinicians must have a voice. Clinicians need the ability to request optimizations, and any changes from that point need to be well communicated. If communication isn’t in place, your clinicians aren’t going to trust the organization to deliver a good experience in the EHR. 4. Address individual pain points. Personalization has expanded in scope, meaning it’s less about how an individual personalization (like a single note template) is going to make an impact and more about finding the solution to a clinician’s specific pain point. 5. Having a strong technical foundation means that your technology functions properly and as promised. If your clinicians don’t feel like your EHR system is reliable or responding quickly enough, your satisfaction just will not improve. 6. Find out how your EHR is contributing to provider burnout. While the EHR is a contributor to burnout, it is not the only contributor. It is a multifactorial problem with varying solutions depending on the individual situation. But there are some key things that stand out in the EHR as clear indicators that someone is burning out. One indicator is that the user is inefficient with their EHR and spending more than five hours a week charting, or they feel like they’re spending too much time in the EHR.

6

https://lnkd.in/gu46i_EN Instead of discussing VSA - Vertical Slice Architecture Discuss team organization. The ability for a few developers to split off within their larger team And work on a feature full-slice. Full-stack, but with teamwork Martin Fowler spells it out well that because of Conway’s Law, this will impact our Architecture And even a backend architecture-only change will typically shine through to the front-end — Monolith vs Microservice? Who cares? If you need a website, build a website If you need a desktop app, build a desktop app If you need a mobile app, build a mobile app Use full-slice development. And you can more easily use other architectures. Like Hexagonal Architecture: The father of microservices — Then designs & patterns should be used

6

1 Comment

I'm going to post a few challenges I have had as a software QA. I should be relaxing today but those that know me, know my brain is always in thirty places at once. To developers: Trust your QA. Y'all don't know how many times I've heard, "Well, it works on my local". I don't care. In the test environment, it breaks. I do my due diligence to investigate why, present evidence in logs and from the database, present recordings, and meticulously document steps of reproduction and results. And even after, I still get, "It works on my local". Get off the dang local. NO ONE knows this software, app, platform, etc. better than the QA because we live it more than anyone else does. To business analysts: I know you all are overwhelmed. But please stop copying and pasting Acceptance Criteria for similar features. It makes a mess for developers and QA because the AC doesn't match or conflicts with not only the developed feature but integration as well. Which brings me to my next point... What QA is: Everyone thinks of us as just testers. I am sick of that. I spend half my time working with the product owner, BAs, and Devs, just making sure the AC reflects the solution needing to be resolved. I can't tell you how many times I have had to address with the BAs conflicts in a story over AC, nonsensical AC, WTF verbiage, etc. So then the developers get confused and we are in long meetings addressing what is messed up. We are here to help you ask the questions you didn't think to ask. We don't just test..... we look at future issues too. We like process improvements. Automation: Stop asking me to automate a one-off feature test that hasn't even gone through UAT yet. It's time wasted. Too many companies do this. Might as well give me the day off for the time it takes to do that because the automation is useless after UAT. Learn the purpose of automation. Product Owner: When I tell you before a demo not to demo certain things because there are still issues, don't shred me after the demo because you decided to demo them anyway and they break. I TOLD YOU not to do that, and why. Data: QA worst nightmare. My last position I would get two test records to test against 20 AC when I first started. I said "Helz no". I built up a catalog of test records I could use, positive and negative, as I progressed because I want multiple records against every AC. Which I often shared with the developers for their unit tests who loved it. QA needs data, all. You give me two records, don't yell at me that coverage isn't thorough. Probably sounded more a rant. But I'm opinionated. Where many QA are not. And these are common issues. Want to improve QA, improve the above.

54

34 Comments

After a third-of-a-century career as a software tester, I'm done. I'm hanging 'em up. I'm retiring. I'm 58 years old and therefore supposed to call this an "early" retirement. I'm not going to let the nominal age at which I can start drawing Social Security benefits define early-vs.-late, though, so I am not going to use that term. I have no grand retirement plans. My horizon is a couple-three months out - I have a few hobbies, and I'm going to install Jira at home and build a backlog of maintenance and improvement projects* with the intent of not being in a particular hurry to start executing. (*both "home" and "personal", if you follow me) You know what, though - this sounds like the first chapter of some kind of adventure novel. "How did I end up getting involved in <something you'd never think I'd do>," something along those lines. So I'll write it out loud here - I'm way past being the protagonist in a thriller story, but if you need someone to be, like, the guy on the other end of the phone who provides you the information or communication assistance so that *you* can undertake the dangerous and exciting adventure, hey, call me up. A career in software testing leaves me with no special adventuring skills or important influential connections, or anything like that, but at least I'll have some spare time, the need for which, in my opinion, isn't emphasized enough in the description of supporting characters in novels and movies.

92

15 Comments

The number of characters that need to change to fix a bug is inversely proportional to the amount of time it takes to find the bug.

3

1 Comment

Can someone explain to me how to get "more experience"? I have 16 years of full stack software engineering experience, mostly in C# .Net 4.2 through 8, including .Net Core 2.x & 3.x, along with WCF, WPF, Winforms, Razor/Blazor, HTML, CSS, and JavaScript, with Entity Framework, Dapper, Linq, SQL Server, MySQL, SQLite, Oracle, NoSql, and other databases. And a few of those years include a combination of Java, PHP, and C/C++. I've also used MarkDown, PowerShell, MVC framework, Swashbuckle/Swagger, Newtonsoft, JSON .Net, TopShelf, and a variety of testing frameworks. But apparently that's not enough. In the past 3 years alone, I've learned ReactJS, TypeScript, Xamarin, XAML, Maui, NodeJS, Terraform, AWS, and Azure, and I'm currently learning Angular, but that doesn't cover the 3-5 years of experience jobs want. I've learned some of these skills through jobs, but also some on my own. Except that as often as companies want us to learn these skills outside of work, they only count professional experience when it comes to hiring. My last interview said I needed more experience, as if the previously mentioned 16 years wasn't enough, and neither was the 15 years of computer repair and networking before that, which their interview questions also covered. So if no one is willing to hire because people don't already have experience, how are we supposed to get the experience? And don't say "open source projects", because they don't count as professional experience. Or do I have to take a 50% pay cut to take an "entry level" job that still requires 3-5 years of experience that I somehow still don't have? People are quick to say that "no one wants to work anymore" when they can't get hired, but they never seem to think "businesses aren't willing to teach anymore" when they refuse to hire people. Why? Is it that we are so conditioned to think that it's always the individuals fault for not living up to expectations, instead of the businesses fault for setting unrealistic and ever changing expectations? So, I ask again, how can I get the required experience when employers refuse to hire unless I already have experience?

2

Great article by Ray Carnes. In my 20+ years in the software industry I've seen companies learn and relearn the basics again and again because each generation of leaders think things are "different this time" because of such and such technology. The frameworks, platforms, underlying technology changes but the principles and foundations are still the same. #softwareengineering

20

3 Comments

I built a simple web app for to-do lists! I actually built quite a few web apps in the last year - namely using Streamlit to avoid learning how to do front-end development. However, I figured it's time to learn it so I can customize the end-user experience a bit more, but also wanted to figure out the following: 1. Learn a Python framework for web apps - Flask, Django, or FastAPI 2. Learn the minimum set of requirements to deploy a web app, end-to-end, so it's accessible publicly 3. Learn how to deploy web apps on Google Cloud Surprisingly, it wasn't very difficult. Some highlights: 1. I chose FastAPI as my framework because it seems to be the cool kid on the block with its super-fast performance, but also because of its swagger docs that are generated automatically. It took about 20 minutes to write the todo web app using FastAPI. 2. I used Jinja templates with FastAPI for my front-end. Super basic, nothing fancy. Took another 10 minutes to figure that out. 3. I deployed the web app locally, just to test it out. Found some bugs and learned more about local deployment of these apps. Spent another 15 minutes here. 4. I created a new project on Google Cloud and used their App Engine cloud service to deploy my web app. Setup took around 10 minutes. Fairly straightforward using their SDK. 5. Embedded the functioning web app to my personal website to make the URLs easier to navigate. Things are always harder until you actually try it. I avoided this for awhile because there was so much technical jargon I just didn't understand. If you've been wanting to learn something new, but afraid to try it out, take this as a sign to go for it! It's easier than you might think! Here's the link to my to-do list web app https://lnkd.in/e_6iztQN. It's accessible to everyone and everyone will see each other's tasks. You can find all the code for this simple app at OpenClinTech: https://lnkd.in/eB_vTbzD. Happy coding! #healthcareonlinkedin #webapp #coding #fastapi #cloudservices

29

3 Comments

Once you hit the #SeniorDev title, your career as a #SoftwareEngineer can go one of three ways. All of these are legitimate decisions and don't ever let anyone tell you otherwise! 1. Management -> This is the oldest and likely the most well-known option. You can head into management and start developing careers of those younger than you. If you love mentoring others and like focusing more on the non-technical management of a project, this is probably the route you want! 2. Staff -> The term "#StaffEngineer" is relatively new to the industry. Many think its title bloat, but I disagree! I think the industry as a whole saw that we needed a new role and we created it. (See my blog on https://iamsellek.com for in-depth info on what a Staff Engineer role looks like!) If you prefer to work in the tec, and you enjoy the bigger-picture aspects of running a codebase/project from the technical side, this might be the route you want! 3. Stay Senior -> Some people look down on this option. Those people are wrong. If you love writing code and doing some light mentoring, there's nothing at all wrong with sticking around as a Senior dev! Some people spend their entire careers here. TLDR? Figure out what you like best about your job. If you prefer mentoring and the non-technical aspects of managing a project, you can head into management. If you want to focus more on the technical side of managing a project and/or a codebase, go with Staff! And if you're happy where you're at, there's nothing wrong sticking with Senior Dev. Now go seize your future! #career #SoftwareEngineer #CareerLadder #future

13

7 Comments

It's so disheartening to see the way predators are circling the developers coming out of bootcamps and other "get rich quick" themed "training programs." As has been said, many of these bootcamps push devs out with little to no real skills into companies who have already agreed to put in a great deal of "on the job training" *because they know you're vulnerable.* When you come out of a bootcamp, you *are not* ready to work and they *know* this. With the advent of narrow AI making this grift even easier to sell, since they can hand these new devs something that makes their programs "work" and shove them out the door, I'm genuinely worried for what's going to happen to the developers coming into this environment. If you're feeling alone and afraid, I do not want to be the person telling you to "stay the path, it's going to be okay." You could have been, or could be vulnerable to be, manipulated by these sort of people. A career in tech is something that's going to take you years to set up, at the very least, and if you can't make that kind of investment, you're going to need to find a way to feed yourself in the meantime because there are sharks circling you *everywhere*. I love this work. I love what I do. I know this is a good career and some people are going to be very happy doing it. If you *aren't* happy doing it, you shouldn't. Do not believe people who say you can get in and make tons of money with very little training, because they're *always* selling you something. Do the work. Put in the hours. Build the tools. Don't believe anyone who tells you it's easy. You can make it, but there aren't shortcuts.

14

4 Comments

There's a lot to consider about selecting EHR and PM software for your practice. Nextech has the most complete and ready-to-go solution. Defiantly Different | All-in-one Solution that fits your practice’s specific needs. Try it and see: https://lnkd.in/e5XTqyzU #dermatology #testimonial #ehr #aesthetic

3

Recently reviewed an enterprise's API style guide. Found this nugget. FYI, style guides are like API documentation - they benefit from a great developer experience. - Standards must be clear and understandable. - Don't assume prior knowledge. Define concepts clearly. - Don't make people guess the intent. Provide examples.

62

1 Comment

I was watching a video that talked about what software developers focus on at different stages of their careers, and one notable take was that "Distinguished Engineers" are focused on the *industry*, and not just their company. This is typically because they've engaged with communities outside of where they're working for many years. They say travel "broadens the mind", and learning from and sharing with developers in a wide range of situations tends to give people a much broader sense of software development. Not everything is Big Tech or Fintech or Medtech or Gametech. Not every product is a smartphone app or a website. Not every product has millions or billions of users. (Proportionally, very few do, in fact; so it's interesting how much we obsess over scale.) Not every code base is 100 million lines. Not every code base is 100,000 lines. Not every system has contemporary hardware to run on. Not every platform can accept updates easily (many embedded systems are ship-once). Not every bug or outage is merely an inconvenience (just ask an airline pilot!) Not every product is about profit. And so on, and so on. Learning about what other people, in other roles, in other organisations, in other industries, in other countries, do will tend to make you a much more well-rounded software professional. It will expand your horizons, as well as create opportunities for the wider developer community to have their horizons expanded from your experiences. In particular, there seems to be fairly wide agreement that a "Distinguished Engineer" has made a notable and lasting contribution to the field and the profession of software development. This seems to be encouraged in organisations who pay it forward; who are happy for their most experienced engineers to "share the love" outside of the walls of their business. When a dev, say, creates a testing framework or writes a book or organises a meet-up on their employer's dime, the employer recognises the long-term value of paying it forward. When employers insist developers engage outside of the business entirely in their own time, and entirely at their own expense, those people will usually end up distinguising some other company. And, of course, if that dev-all-the-way-to-the-top career track isn't on offer, then ditto :-)

40

Most off-the-shelf electronic health record solutions (EHRs) don't work in lockstep with your organization, but that doesn’t mean you have to settle for less. Overhaul your workflows with a personalized acute care EHR that adapts to your needs and preferences. https://lnkd.in/e-zGKeFF

12

Three years ago I wrote a post arguing for better tools for developers working with FHIR. I’d just come out of a contract with a major US health insurance company. It was my introduction to FHIR and it was a frustrating introduction. The tools I expected to find as a developer did not exist. FHIR is 10 years old. That’s almost antiquated in tech years, which tend to run at a faster pace than dog years. Look at what ChatGPT has achieved in only 2 years. Love it or hate it, that’s real progress. I can’t help but think FHIR should be further along than it is. Not the standard itself, which moves at a slow pace for a reason, but the surrounding ecosystem. - We should have better tools. - We should have more uniformity on best practices. - We should have documented architectures for building big FHIR solutions. - We should have better conversion to and from FHIR. We don’t have any of these things. In that three year old post I was asking for a desktop client for accessing FHIR APIs. A small ask, from a frustrated developer. It was 6 months later that I built the first version of the Vanya FHIR Viewer. Vanya then: https://lnkd.in/ejV-kBhK Vanya today: https://lnkd.in/eeY22em2 Here’s that first post from 3 years ago: https://lnkd.in/dyDPWeMw ~~~ Download Vanya for Windows and Mac: https://lnkd.in/e-7ZgCKV

76

11 Comments

Do you debounce? Working on a project, and someone pointed out that a certain API wasn't debounced from the front-end. It was firing - after every. Keystroke. That adds up. To estimate how many requests are saved by adding a 300ms debounce to a search that initially fires on every keystroke, we need to look at a few factors: 1. Typing Speed: Average typing speed is about 200–300 characters per minute (cpm), which translates to approximately 3–5 keystrokes per second. 2. Without Debounce (Every Keystroke): Each keystroke triggers an API request. 3. With Debounce (300ms): A 300ms debounce means that only one request is sent if there’s a 300ms pause after the last keystroke. Let's calculate the potential reduction: 1. Estimate of Requests Without Debounce: - If a user types at 3–5 keystrokes per second, they might trigger 3–5 requests per second without a debounce. 2. Estimate of Requests With 300ms Debounce: - With a 300ms debounce, if the user pauses briefly between typing bursts (common in real usage), the debounce will typically group each sequence of keystrokes into 1 or 2 requests per second, significantly lowering the request count. 3. Reduction Factor: - By adding the 300ms debounce, you’d reduce requests by around 70–80%, depending on typing patterns. Example Calculation for a 10-second search session: - Without Debounce: At 3–5 requests per second, this would be 30–50 requests. - With 300ms Debounce: This would likely reduce to around 6–10 requests. So, for each 10-second session, you’re saving approximately 24–40 requests.

5

1 Comment

15 years ago I contributed 2 chapters to Clean Code(https://a.co/d/5guLEzq), both related to threading. A second edition is coming out, and it occurs to me that while I don't do much direct threaded/parallel programming anymore, many issues around parallel execution still occur in systems where the stack "takes care" of all those details. Here's a collection of recent examples. This is a draft of a personal blog. The second version will go on the Industrial Logic, Inc. website. The final version will likely be added as an update for the 2nd Edition of Clean Code. Comments welcome, enjoy! https://lnkd.in/gWCUkK6V

9