DEV Community: Saviour Barry

HOW I AUTOMATED MY DAILY WORKFLOW USING AI (AND GOT MY TIME BACK)

Saviour Barry — Sat, 11 Apr 2026 20:48:50 +0000

Every morning, I used to wake up feeling like a productivity machine.
“This is the day. I’m getting everything done.”
Then I open my laptop…
One email turns into ten.
One quick reply turns into a full conversation my notes look like something written during an earthquake and somehow, two hours are gone… just like that.

At some point, I had to ask myself a very serious question:

Why am I doing all this manually like it’s 1997? No offence Millennials.

So I tried something new I started using AI to handle the boring parts of my day.

And no, I didn’t suddenly become a tech genius overnight. I just stopped suffering unnecessarily

It started with emails, the ultimate productivity trap. You open one message, and before you know it, you’re mentally involved in five different conversations you never planned to have. Replying used to take forever because I’d overthink every sentence like I was writing a formal letter to the president.

Now? I let AI do the first draft.

Using tools like Gmail with AI features, I get quick summaries and suggested replies. I just tweak them a bit and hit send. What used to take an hour now takes minutes, and I no longer feel emotionally drained by my inbox.

Then came my notes… which, honestly, were a disaster.

I used to write things like:
“important idea!!! check later maybe???”

Later would come, and I’d have no idea what I meant. Was it an idea? A reminder? A cry for help?

Now, with Notion AI, my notes actually make sense. It organizes them, cleans them up, and somehow turns my chaotic thoughts into something readable. It’s like having a very patient friend who understands my confusion and fixes it quietly.

Planning my day used to be another struggle.

I’d sit there thinking:
“Okay… what should I do now?”

Then I’d open another app. Then another tab. Then suddenly I’m watching random videos I didn’t plan to watch.

But with tools like Google Calendar, my tasks now have structure. I don’t have to guess what to do next it’s already laid out for me. It feels less like chaos and more like I actually have control over my time.

The funny thing is, my system isn’t complicated at all.

I just:

Dump everything I need to do in one place
Let AI organize it
Follow the plan

That’s it.

No stress. No overthinking. No “I’ll figure it out later” (because later never comes).

Now, let’s be honest AI is not perfect.

Sometimes it gives weird suggestions. Sometimes it confidently says something that makes no sense. And yes, you still have to use your brain (sigh....unfortunately).

But that’s fine.

Because the goal isn’t to replace yourself. The goal is to stop wasting energy on repetitive tasks that don’t need your full attention.

As a student and developer, this honestly changed everything for me.

I used to spend so much time trying to organize my life that I barely had time to actually live it or build anything meaningful.

Now, I focus more on:

Learning new things
Working on projects
Getting real work done

And less on:

Rearranging tasks
Rewriting notes
Fighting with my inbox

At the end of the day, AI didn’t magically make me more productive,
It just removed the small, annoying obstacles that kept slowing me down.

And once those were gone?
Everything else became a lot easier.

AiAssist #WorkflowAutomation

Wait, Your AI Can See That? Lol....Lets Talk About Multimodal Tech

Saviour Barry — Tue, 10 Mar 2026 21:57:50 +0000

Have you ever noticed how humans learn? We don’t just read text. We see colors, hear sounds, and feel textures. For a long time, AI was "unimodal" meaning it could only do one thing at a time, like analyze text or recognize an image.

But the game has changed. Enter Multimodal AI.

What Exactly is Multimodal AI you ask?
In simple terms, Multimodal AI is a type of machine learning that can process and "understand" different types of data (modalities) simultaneously.

Think of it like this:

Unimodal AI: A librarian who only reads books, cool but sorta boring.

Multimodal AI: A person watching a movie who understands the dialogue (text), the soundtrack (audio), and the acting (visuals) all at once sounds more cooler right.

Why Should Developers Care?
If you're building apps in 2026, you're no longer limited to text-in/text-out (you saw that right dudes and dudettes) . Multimodal models like Google’s Gemini or Open AI’s GPT-4o allow you to build features that were nearly impossible a few years ago:

Visual Q&A: Upload a photo of a broken car engine and ask the AI, "How do I fix this?"

Video Summarization: Send a 10-minute lecture and get a bulleted summary of the key points.

Real-time Accessibility: Convert images and surroundings into descriptive audio for the visually impaired.

How Does it Work? (The "Under the Hood" Lite Version)
You don't need a PhD to understand the basics. Most multimodal systems follow a three-step process:

Encoders: Each input (image, text, audio) has a "specialist" that turns it into numbers.

Fusion: This is the "magic" step where the AI aligns these numbers. It learns that the word "Golden Retriever" in text relates to the pixels of a fluffy yellow dog in an image.

The Brain (LLM): A large language model takes that combined information and gives you a human-like response.

Multimodal AI is making our applications feel more "human" and context-aware. Whether you're building a tool for students to analyze geological maps or a fitness app that "sees" your form via camera, the possibilities are endless.

What are your thoughts about Multimodal AI? Let me know in the comments.

AI #MachineLearning #Python #Beginners

AI Is Taking Over Every Sector : But As an African, do i Still Need to Learn How to Code

Saviour Barry — Wed, 04 Mar 2026 12:10:49 +0000

Okay so around 2021 I was so convinced that I had this whole path thing all figured out going into the tech industry as a newbie, you know learn web development and get paid fat stacks of cash as a google employee 'kinda stuff lol',

fast forward to 2025 as a frontend developer, I wasn't so sure of what I wanted again at some point I was like what's the need to even learning to code or forward my code knowledge because the "Artificial Intelligence" as the Germans would say "za' AI" is rapidly transforming every major sector of the global economy. From banking systems and healthcare diagnostics to fashion recommendations and government automation,

AI infrastructure is quietly becoming the backbone of modern digital systems. Major platforms powered by organizations like OpenAI, cloud ecosystems developed by Microsoft, and machine learning frameworks engineered by Google are shaping how businesses operate and how users interact with technology.

AI didn't just level the playing field, it came and uprooted the four legs of the table and flung the chairs outside

As this transformation increases every day, I then asked my self: if AI is doing so much of the heavy lifting, do individuals especially Africans still need to learn how to code or more importantly do I need to forward my coding knowledge?

Now I've done a lot of research and arguing with fellow developers, and eventually The answer is yes, as an Africans we should be encouraged to learn coding, but as well learn it fast, I feel in this era we shouldn't waste time mastering languages, but instead learn how they work from variables, classes, functions to Db queries, to simply put learn how to craft scalable business logics with the languages especially now more than ever.

As Africans I think we should dwell more on solving indigenous problems with code, and look for opportunities opening around other fields and apply domain knowledge to it.

The thing is AI can write code, but AI cannot create new job roles. to create roles require a different level of intelligence that is far beyond the reach of AI, mind you even the AI models of 2026 are not truly intelligent, they are just good at probability math's (although I'm not so good at probability math's my self lol).

All I'm trying to say is designing your own role and presenting it to a company takes real human intelligence and we already have this. I mean our ancestors were problem solvers, some of them literally created problems and offered solutions for it lol, not to sound mischievous here but that's how the real world operates

AI is powerful, but it is not magic. It does not operate independently of systems. It runs on servers, APIs, databases, authentication frameworks, and frontend interfaces. It requires infrastructure to be deployed effectively, and that infrastructure must be built, maintained, secured, and customized by developers. AI can generate responses, analyze data, and automate workflows, but it cannot independently architect complete digital ecosystems tailored to local realities. Developers build systems; AI enhances them. For Africa, where digital transformation is still unfolding across many regions, system builders are not optional they are essential.

In an age defined by artificial intelligence, coding remains a human advantage and for Africa, it may be one of the most important investments of this generation.