Essence Solusoft

29/09/2026

Everyone is asking: will Jev replace AI models like ChatGPT? 🤔

Not really. Jev is a different kind of model. Instead of writing paragraphs, it makes quick decisions for software, such as "Is this message urgent?", "Which team should handle this?" or "Is this action safe?", and it does it fast and cheaply.

Think of it this way:

⚡ Jev = quick instinct decisions

🧠 LLMs = deep thinking, writing, and reasoning

The best AI apps will likely use both together.

We're following this closely at Essence Solusoft, where we build Rails apps and AI-powered solutions for businesses.

What do you think: hype or the next big shift? 👇

24/09/2026

Why Should Customers Wait for Work Rails Can Do Later?

A customer completes checkout. The payment succeeds - but the page keeps spinning while the app sends a receipt, generates a PDF, and calls another service.

Under heavy traffic, that wait can become a timeout, a second click, or an abandoned cart.

Rails background jobs solve the immediate problem: complete the critical work, return a confirmation, and let workers handle tasks that can run later. But a fast response means little if the queued work fails or runs twice.

That’s why reliable job architecture matters. Jobs need clear responsibilities, appropriate queue priorities, safe retries, and monitoring. Order-related jobs must be enqueued at the right time, too: a receipt job that runs before the order’s database transaction commits can fail - or send a receipt for an order that was rolled back.

At Essence Solusoft, we choose the queue backend to fit the application. A new Rails 8 project may suit Solid Queue; an existing app with a stable Sidekiq setup rarely needs a migration just for the sake of switching tools.

The goal is simple: customers get a prompt response, and the work they depend on still gets done reliably.

📖 Read the full blog: https://essencesolusoft.com/rails-background-jobs-why-customers-shouldnt-wait-for-work-that-can-run-later/

23/09/2026

🏗️ Good architecture shouldn’t depend on pull request comments and tribal knowledge.

At Essence Solusoft, we’ve started standardizing the architecture of our Rails projects using ArchSpec, an open-source gem that turns architecture rules into automated checks.

⚙️ Getting started took minutes:

1️⃣ Add the gem and run bundle install
2️⃣ bundle exec archspec init to create the rules file (Archspec.rb)
3️⃣ bundle exec archspec check to scan the project

🔍 Our first scan of a live project found 3 violations:

❌ Two models depending on controller-layer code
❌ One service reaching directly into params

Nothing was broken, but these boundaries would have kept getting blurrier over time. Now they’re visible, and each error points to the exact file and line.

💡 What we like about it:

✅ Rules are written in plain Ruby
✅ It’s static analysis, so it never boots the app and no AI is involved
✅ Existing violations can be recorded in a todo file and fixed step by step
✅ It can run in CI, so every change is checked, whether a person or an AI assistant wrote it

Clean architecture isn’t a one-time refactor. It’s a habit you enforce automatically.

👏 Credit to the ArchSpec creator for a very useful tool.

https://github.com/crmne/archspec

18/09/2026

🙏 Some goodbyes are just promises to meet again.

Bappa, thank you for filling our workplace with blessings, positivity, and togetherness. ❤️
Until we welcome you again… Ganpati Bappa Morya! 🐘🌺

18/09/2026

🎉 Happy Birthday, Ashok Damaniya! 🎂

Wishing you a day filled with happiness, laughter, and wonderful moments. ❤️
Here’s to another year of success and great memories!

Cheers to you! 🥳✨

15/09/2026

Do you really need Redis for background jobs anymore? 🤔

For years, Redis + Sidekiq has been the standard way to run background jobs in Rails. Fast, reliable, proven.

But Rails 8 now ships with Solid Queue as the default a background job backend that uses your existing database instead of Redis. No extra service to install, deploy, or babysit.

Redis still wins on raw speed at high volume. But for a lot of apps, Solid Queue means one less piece of infrastructure to worry about and that simplicity is huge.

Heads up though: since Solid Queue stores jobs (even finished ones) in your database, apps doing a LOT of background processing will see their database grow faster. Worth setting up regular cleanup for that. 🧹

Which one are you using for background jobs these days?

14/09/2026

🙏✨ Celebrating Ganpati Bappa together as Team Essence!

Ganpati Bappa Morya! 🌺

14/09/2026

What Breaks When an AI-Generated MVP Goes to Production?

An AI-generated MVP can look impressive in a demo. Production is where the real test begins.

Real users bring traffic spikes, messy data, rate limits, unpredictable model outputs, security risks, slow responses, duplicate jobs, and integrations that don’t behave the way a demo environment does.

The biggest problem is rarely the model itself. It’s everything around the model.

AI MVPs need solid application architecture, reliable data, rate limiting, background processing, fallbacks, observability, security, and a clear way to measure whether the AI is actually delivering value.

At Essence Solusoft, we approach AI MVP development as Ruby on Rails product engineering first, AI second. We start with one clear business problem and metric, validate the data, integrate the right AI approach, and gradually roll it out with monitoring and real user feedback.

Because the gap between a working AI demo and a production-ready product is where most AI MVPs break.

📖 Read the full blog: https://essencesolusoft.com/ai-mvp-development-what-breaks-in-production/

Most AI MVPs work great in a demo and fall apart in production. Here’s what actually breaks, and how we build AI MVPs on Rails that don’t.