DEV Community: Flora Brandão

Choosing a tech stack for interactive AI apps 🤖

Flora Brandão — Tue, 22 Sep 2026 09:02:54 +0000

Coding agents completely changed how teams select a tech stack. The old grind of choosing tools based solely on how fast you can write code causes real headaches when you try to actually run interactive AI applications in production.

Here is how to approach stack selection today:

The problem:
Evaluating tools based on writing speed ignores operational reality.

The fix:
Focus on whether you can reliably run the app rather than just write it. Take a tour of 8 key ecosystems and eliminate bad options before making a choice.

Why it matters: Selecting a resilient ecosystem ensures your interactive AI app stays manageable as coding agents evolve.

Check out the full technical write-up to explore the ecosystem breakdown and selection criteria:

Which stack would you pick for an interactive AI app today? - Upsun Developer

Coding agents changed how new teams pick a tech stack. A tour of 8 ecosystems for interactive AI apps, and why "can we run it" beats "can we write it".

developer.upsun.com

Offloading heavy jobs: running tasks with on-demand containers ⚡

Flora Brandão — Thu, 10 Sep 2026 09:14:22 +0000

Running long, resource-hungry jobs directly in standard crons is a constant grind. Database imports and search reindexing can easily choke your environment and hog shared resources.

Here is how to offload heavy background work:

The problem: Long-running jobs strain resources and slow down standard cron execution.
The fix: Move heavy processes into on-demand Upsun task containers that are authorized and triggered by a lightweight cron.
Why it matters: Your core app stays fast, while heavy tasks get dedicated environment resources only when fired.

Check out the full technical write-up to see how to define, authorize, and trigger task containers for your stack:

Offload Heavy Jobs to Upsun Task Containers, Triggered by Cron - Upsun Developer

Move long, resource-hungry jobs out of your crons and into on-demand Upsun task containers. Learn how to define a task, authorize your app to trigger it, and fire it from a lightweight cron for database imports or search reindexing.

developer.upsun.com

Migrating apps without rewriting code 🚀

Flora Brandão — Tue, 08 Sep 2026 14:44:41 +0000

Application migrations often stall because rebuilding deployment pipelines and environment setups takes months of tedious work. Your application code is rarely the problem, but rebuilding the surrounding infrastructure is a massive headache.

Here is how you can simplify the process:

Declare your runtimes, services, and routes once in a single configuration file stored in your repository.
Push a Git branch to automatically launch an isolated environment for testing.
Sync real production data into your branch to validate changes safely before merging.
Keep your existing application logic, CLI tools, and CI pipelines intact without refactoring.

Testing happens on a dedicated branch, so you never have to test in production or risk live site downtime.

Check out the full article see the step-by-step workflow:

Migrate your app to Upsun without a rewrite | Upsun

Migrating to Upsun doesn't mean rewriting your app. Add one config file, push a branch, test against real production data, and merge only when it works.

upsun.com

The real cost of cloud lock-In ☁️

Flora Brandão — Mon, 07 Sep 2026 14:40:51 +0000

Cloud dependency rarely starts with a massive strategic decision. It accumulates quietly through thousands of reasonable technical choices until switching becomes a multi-million dollar constraint.

The core issues:

Egress fees consume 10 to 15 percent of total monthly cloud budgets
AI workloads accelerate lock-in due to provider-specific GPU hardware and proprietary tooling
Enterprise migration costs range from $500,000 to millions when leaving a vendor

The fix:

Standardize on portable environment definitions at the delivery layer
Build pipelines using open standards rather than vendor-specific APIs
Maintain true architectural optionality before proprietary dependencies stack up

Why it matters:
Portability gives your team real leverage during contract renewals and prevents pricing changes or service deprecations from disrupting your roadmap.

Check out the full technical artcile for the complete breakdown and migration reference architecture:

The real cost of hyperscaler dependency | Upsun

AWS, Azure and GCP hold 60% of cloud infrastructure. See how hyperscaler dependency erodes your leverage — and what real portability requires

upsun.com

Upsun is recognized in the Gartner® Magic Quadrant™ for Cloud-Native Application Platforms

Flora Brandão — Mon, 07 Sep 2026 10:30:59 +0000

We have done it again. 🎉

For the third consecutive year, Upsun is recognized in the Gartner® Magic Quadrant™ for Cloud-Native Application Platforms.

We believe this recognition reflects the platform you rely on every day to ship software, manage infrastructure, and keep pace with the demands of modern development, now including AI-powered workflows and agentic software delivery.

Unlock critical insights to future-proof your cloud strategy with the 2026 Gartner® Magic Quadrant™ for Cloud Application Platforms report:

Upsun acknowledged for its Ability to Execute and Completeness of Vision | Upsun

Upsun acknowledged for its Ability to Execute and Completeness of Vision for third consecutive year in Gartner® Magic Quadrant™ for Cloud-Native Application Platforms.

upsun.com

How exposed is your stack to cloud lock-In? ☁️

Flora Brandão — Fri, 04 Sep 2026 14:14:22 +0000

Maintaining separate CI/CD pipelines for every cloud provider and manually gathering audit evidence across consoles is a painful engineering grind.

The problem:
Cloud dependency concentrates in four places and slows down every change: - Applications tightly bound to provider-specific services

High friction and long timelines for planned migrations
Inconsistent delivery workflows across different teams
Non-portable governance and audit controlsThe Fix:

Consolidate at the delivery layer to fix the root cause:

Store one portable application definition in Git
Declare policy as code at the platform layer
Use a single delivery workflow for every target cloud

Why it matters:
You lower switching barriers and make provider choice a downstream decision without rewriting your whole architecture.

Read the full article to score your own cloud exposure:

Multicloud exposure assessment for IT leaders | Upsun

Assess how exposed you are to cloud lock-in in about 15 minutes. Score four areas of multicloud dependency and see which risks to address first

upsun.com

Demystifying SSH authentication: how certificates and proxies Work 🔑

Flora Brandão — Fri, 04 Sep 2026 09:37:19 +0000

Managing secure access and permissions across environments can quickly become a tedious engineering headache. Understanding the exact flow of your SSH connection is essential for keeping systems secure without losing visibility.

Here is how SSH authentication functions under the hood:

The problem: SSH connections can feel like a black box when dealing with certificate provisioning, trust, and multi-factor authentication.
The fix: Standardize access using an SSH proxy, CA trust validation, certificate decoding, and structured permission extensions.
Why it matters: You get clear insight into your access checks, permissions, and final connection hops to keep your infrastructure secure.

Check out the full technical write-up to see how the entire SSH authentication flow works step by step:

How SSH authentication works on Upsun - Upsun Developer

Decode the SSH certificate the Upsun CLI provisions for you, then follow it through the SSH proxy: CA trust, revocation, access documents, and MFA.

developer.upsun.com

Stop skipping stages in your AI adoption 🤖

Flora Brandão — Thu, 03 Sep 2026 15:05:23 +0000

Imagine opening your laptop on Monday morning to 53 changed files and over 2000 diffs dumped by unsupervised coding agents.

Here is why jumping straight to fully autonomous AI breaks team workflows:

The problem: Individual AI productivity outruns team readiness, leaving developers feeling ambushed and unable to review code properly.
The solution: Adopt AI sequentially through an eight-stage maturity model instead of skipping straight to autonomous factories.
Why it matters: Trust must be earned workflow by workflow, ensuring review habits and governance are built before automation scales.

Check out the article to learn how to navigate each stage of AI maturity:

Don't build the autonomous AI factory first | Upsun

Skipping stages in AI adoption creates security and technical debt for engineering teams. Learn the eight stages to safely build an autonomous AI factory

upsun.com

How to choose a code index for your coding agent 💻

Flora Brandão — Thu, 03 Sep 2026 09:53:28 +0000

Finding the right context index for a coding agent across Go, TypeScript, Swift, and SQL codebases is a real engineering challenge. Without proper context coverage and retrieval quality, your agent struggles to deliver accurate results.

Here is what was learned testing Graft and roam-code on a real project:

The problem: Agents need the right balance of context coverage, retrieval quality, and change analysis to be useful.
The fix: Graft delivers more direct and focused results, while roam-code adds helpful change analysis features with specific caveats.
Why it matters: Proper index coverage changes the final recommendation and ensures your agent provides actionable assistance.

Check out the full article to see the complete evaluation and test results:

How to choose a code index for your coding agent - Upsun Developer

Compare Graft and roam-code on a real Go, TypeScript, Swift, and SQL codebase to evaluate context coverage, retrieval quality, and change analysis in practice.

developer.upsun.com

Why managing multicloud at the provider layer is breaking your governance ☁️

Flora Brandão — Wed, 02 Sep 2026 14:04:33 +0000

Managing separate compliance controls, access policies, and billing consoles for AWS, Azure, and GCP manually is an absolute operational grind.

The problem:

Infrastructure spread across multiple public clouds through acquisitions, team choices, and disaster recovery needs
Traditional governance frameworks stop at the boundary of each provider console
Managing policies per environment leads to configuration drift, security breaches, and scaling headaches

The fix:

Move governance up to the delivery layer above individual cloud providers
Define access controls, policies, and environment configurations once in version controlled code
Enforce guardrails automatically across all environments instead of jumping between consoles

Why it matters:

Eliminates manual audit prep and fragmented cost tracking
Maintains workload portability without rebuilding security models
Reduces vendor lock in and keeps negotiating leverage in your hands

Check out the full article to learn more about setting up cross provider governance guardrails:

Multicloud governance: policy across providers | Upsun

Multicloud is the reality most IT teams already have. See why governance, not infrastructure, now decides resilience, compliance, and control

upsun.com

AI tools forgot that software is a team sport 🤝

Flora Brandão — Tue, 01 Sep 2026 14:04:36 +0000

Individual AI tools make coding faster for one person, but they do not solve the work that happens between people. More AI-generated code just adds drag to the review queue.

The problem: AI speeds up solo code creation, but pull request reviews, discussions, and architecture debates still cause bottlenecks.
The fix: Focus on structured workflows rather than standalone agents. Run agents in isolated sandboxes with explicit human approval gates for critical decisions.
Why it matters: It brings product managers, designers, and security teams directly into the process. The review load spreads out instead of piling up on senior engineers.

Check out the full article to see how teams are testing agent-run reviews and human-in-the-loop workflows:

Software is a team sport. Most AI tools forgot that.

AI made writing code fast. Upsun product manager Kateryna Dvornichenko on why team collaboration, not individual speed, is the real test for AI agents

upsun.com

Why the fastest teams standardize first 🏎️

Flora Brandão — Tue, 01 Sep 2026 09:56:39 +0000

Developers spend 40 to 60 percent of their time on non-coding, non-strategic toil like broken staging environments and environment drift. Ad hoc speed feels fast, but variability throttles overall throughput.

Here is how standardizing the platform layer changes the daily workflow:

Fix the core problem: Inconsistent deployment scripts and undocumented configurations create structural drag.
Apply the platform fix: Automate standard paths for deployments, environment definitions, and security gates.
Free up senior engineers: Stop turning senior devs into accidental DevOps leads so they can focus on core product decisions.

Standardization acts like better brakes on a car. It provides the control needed to move faster safely.

Check out the full article to read the complete breakdown:

Why high-velocity teams standardize first | Upsun

High-velocity teams standardize the path to production first. See why repeatable velocity beats ad hoc speed, and how it returns time to developers.

upsun.com