Hello 👋 I'm Aravind HU

Exploring the Future of Software Engineering

Evolutionary Architecture • Platform Engineering • Agentic AI • Developer Experience

I've been building software since 2009—but my real passion isn't just building software. It's understanding how software engineering itself evolves.

Aravind HU
2009

Building Software Since

From application code to the platforms and systems that produce it

17+

Years of Experience

Healthcare, construction, SaaS—architecting platforms that scale

23

Insights Shared

Real-world lessons on architecture, platforms & developer experience

How I See the Engineering System

Everything I work on sits somewhere on this map. The business pulls architecture; architecture shapes the platform; the platform shapes what engineers can actually do. 

                    ┌──────────────────────────────┐
                    │    Business & Customer       │
                    └──────────────┬───────────────┘
                                   │
                                   ▼
                   ┌────────────────────────────────┐
                   │   Evolutionary Architecture    │
                   └───────────────┬────────────────┘
                                   │
          ┌────────────────────────┼────────────────────────┐
          ▼                        ▼                        ▼
 ┌─────────────────┐      ┌─────────────────┐      ┌─────────────────┐
 │ Platform        │      │ Developer       │      │ Engineering     │
 │ Engineering     │      │ Experience      │      │ Effectiveness   │
 └────────┬────────┘      └────────┬────────┘      └────────┬────────┘
          │                        │                        │
          └──────────────┬─────────┴─────────────┬──────────┘
                         ▼                       ▼
              ┌─────────────────────────────────────────────┐
              │      AI-Native Engineering Platform         │
              │   Agentic Workflows • Intelligent Systems   │
              └──────────────────────┬──────────────────────┘
                                     ▼
                      Better Software. Better Teams.

What I'm Exploring

🧠 Agentic AI for Software Engineering ⚙️ Intelligent Internal Developer Platforms 🏗 Evolutionary Architecture ☁️ Cloud Native Platforms 🚀 Platform Engineering 🌉 Engineering Effectiveness 🤖 AI-native Engineering Workflows 📚 Engineering Knowledge Systems 🔄 Continuous Architecture 🧩 Developer Experience

The Thread I'm Following: Agentic AI in Engineering

My journey moved through evolutionary architecture, cloud native systems, platform engineering and internal developer platforms into AI-native software engineering. Developer experience and engineering effectiveness are not stops along that route—they are the lens I judge all of it through, and still the thing I research most.

Today I'm exploring how Agentic AI becomes part of engineering itself. Not simply writing code—but helping engineers design, build, test, secure, deploy and evolve software.

At Philips I've been a Senior Software Architect since 2019, working on healthcare platforms and digital solutions from Bengaluru, India. That tenure is where the thread actually plays out—in a regulated environment, where anything you automate still has to be defensible. It has moved in three stages:

  • 01 AI-assisted development. Driving adoption of Claude and GitHub Copilot across teams—a practical productivity multiplier, and the first honest look at what engineers will and won't delegate.
  • 02 AI-native platform thinking. Asking what an internal developer platform should become when AI is not a plugin but a participant—and what the golden paths need to look like for it.
  • 03 Agentic workflows, AI harness engineering & RAG. Where I am now, and hands-on for the last couple of years—building the harness around the model rather than just prompting it: tool boundaries, context strategy, evaluation and guardrails. Alongside that, RAG systems so engineering knowledge stays retrievable instead of re-derived, by engineers and agents alike.

Deliberately multi-model. I work across Claude, GitHub Copilot, Antigravity and local runtimes like Jan AI. Harnesses outlive models, so the useful skill is building one that isn't captive to a single vendor—and knowing when the answer is a local model rather than a hosted API.

The Foundation It's Built On

None of the AI work is interesting without a platform underneath it. What I've delivered:

  • Internal Developer Platform at Philips—standardized, automated build and deploy workflows built on Backstage
  • InnerSource practices institutionalized—amplifying reuse and reinforcing shared ownership across teams
  • Shift-left quality practices that nearly halved release cycles while improving reliability
  • Golden paths that enable self-service without compromising governance or compliance

Reference Implementation: A Self-Driving Repository

I'd rather build the idea than describe it. So the agentic thread has a public artifact—a .NET 10 repository where the only human input is a GitHub issue.

  • Agents write the failing test, implement it, review the pull request, and merge it—no human reads the diff.
  • CI mechanically verifies test-first development: it checks out the test commit and requires the suite to fail there. A claim of TDD costs nothing; a red commit doesn't.
  • Agents cannot edit the gates that judge them—which is what makes merging without review defensible at all.
  • A Karpathy-style LLM Wiki knowledge layer that is maintained after every merge, not regenerated on demand—knowledge should compound.
calculator-with-llm-wiki Read: The Self-Driving Repository: Letting Agents Merge Their Own Code

The Lens: Developer Experience

Platforms and AI are means. Developer experience is how I decide whether either one actually worked—and it is the area I research most consistently.

  • Cognitive load is the real cost. Most "productivity" problems are load problems wearing a disguise—too much to hold in your head to make one safe change.
  • Feedback loop length beats tool count. How long until an engineer finds out they were wrong is the number I keep coming back to.
  • Measuring DevEx without gaming it. Reduce it to velocity and you get velocity. The interesting research is in measures that survive being targets.
  • And now: experience for a non-human user. Agents read your docs, your golden paths and your error messages too—and they are far less forgiving of ambiguity than engineers are.

Current Research

The questions I'm actively working on. I don't have finished answers to any of them yet—that's rather the point.

  • How do engineering organizations evolve in the age of AI?
  • What should Internal Developer Platforms look like when AI agents become first-class users?
  • Can platforms actively guide engineering decisions rather than simply automate tasks?
  • How does Evolutionary Architecture change when software is built collaboratively by humans and AI?
  • What does developer experience mean when some of your developers are agents—and how do you measure it without reducing it to velocity?
  • What does an Engineering System look like in 2030?

Areas of Interest

Architecture

  • Evolutionary Design
  • Distributed Systems
  • Domain Driven Design
  • Cloud Native

Platform Engineering

  • Internal Developer Platforms
  • Golden Paths
  • Self-Service
  • Engineering Platforms

Artificial Intelligence

  • Agentic AI
  • AI-native Engineering
  • Intelligent Workflows
  • Engineering Knowledge

Developer Experience

  • Cognitive Load
  • Feedback Loops
  • Engineering Effectiveness
  • Measuring DevEx

Philosophy

  • Technology changes. Engineering principles endure.
  • Platforms should reduce cognitive load.
  • Architecture should evolve continuously.
  • AI should augment engineers—not replace them.
  • The best engineering organizations continuously improve the systems that build software.

Core Expertise

Agentic AI AI Harness Engineering RAG Systems AI-Native Engineering Evolutionary Architecture Platform Engineering Developer Experience Internal Developer Platforms Engineering Effectiveness Backstage InnerSource Cloud Native Golden Paths CI/CD

Currently

Senior Software Architect

Philips · 2019 – Present

Healthcare Platforms & Digital Solutions · Bengaluru, India

Architecting internal developer platforms for regulated healthcare environments—and exploring what those platforms become when AI agents are first-class users.

You'll Usually Find Me

  • 💡 Building engineering platforms
  • 🤖 Experimenting with Agentic AI
  • 📝 Writing about software engineering
  • 🎤 Speaking at engineering communities
  • 🧪 Building reference implementations
  • 📖 Learning something new every week

Let's Connect

Open to thoughtful conversations on evolutionary architecture, platform engineering, agentic AI, developer experience, or scaling delivery in regulated environments.

Email Me

Exploring the Future of Software Engineering.