dead-weight athulsreekumar

I turn ideas into intelligent products — and the cloud platforms that keep them running.

Most people do either the AI or the infrastructure. I do both:
I design the models' workflows, wire them into real products, and build the pipelines,
security and cloud underneath so they hold up in production.

◆ What I build

🧠 AI Systems

LLM-powered applications, autonomous agents and retrieval pipelines that work on real data — not demos.

☁️ Cloud Platforms

Complete Azure environments as code: networking, containers, secrets, monitoring — secure by default.

⚡ Products

Fullstack web and mobile apps, shipped through pipelines that test, build and release themselves.

◆ AI — what I work on

Explore

Area	What that means
LLM applications	Production apps built on large language models — prompt design, structured outputs, tool/function calling, streaming UX
AI agents	Multi-step agents that plan, call tools and APIs, and take actions inside real workflows
RAG & knowledge	Embeddings, vector search and retrieval pipelines that ground answers in private data
AI automation	Replacing manual, repetitive work with AI-driven pipelines and integrations
Evaluation & guardrails	Testing model output, catching failures, keeping cost and latency in check
LLMOps	Deploying, versioning, monitoring and securing AI services in the cloud

How an AI feature goes from idea to production

flowchart LR
    I([💡 Idea]) --> P[Prototype<br/>prompts + tools]
    P --> R[Ground it<br/>RAG / data]
    R --> E{Evaluate}
    E -- not yet --> P
    E -- good --> S[Ship as API<br/>container]
    S --> C[CI/CD → Azure]
    C --> M[Monitor<br/>quality · cost · latency]
    M -. feedback .-> P

◆ Cloud & DevOps

Explore

Full Azure infrastructure in Terraform — VNets, containers, registries, Key Vault, App Insights
CI/CD with GitHub Actions and Azure DevOps — build, test, scan, deploy on every push
Automated iOS & Android releases straight from the pipeline
Secret management end-to-end across infra, pipelines and containers
Microservices running production mobile apps at scale
Azure Cloud Professional

◆ Open source

Fixing real bugs in the tools developers use every day.

Merged

Project	PR	What it fixed
Starlight (Astro)	withastro/starlight#4309	Sidebar directories named like `Object.prototype` properties (`constructor`, `toString`)
MDN Web Docs	mdn/content#46321	`referrerpolicy` / `noreferrer` docs now cover downloads

In review

Project	PR	What it fixes
Biome	biomejs/biome#12734	CSS strings quoted with HTML entities (`"`) in `style` attributes
Biome	biomejs/biome#12740	`useValidAriaValues` false positive on Astro template literals
oxc	oxc-project/oxc#28405	Lexer panic on files with more than 1024 diagnostics
oxc	oxc-project/oxc#28406	Unterminated block comments recorded with the wrong kind
PhotoCraft	storytold/photocraft#10123	Levels histogram stops at 300 px
PhotoCraft	storytold/photocraft#10124	Histogram panel graph doesn't grow with the panel
PhotoCraft	storytold/photocraft#10279	Curves graph doesn't fill the panel height

◆ Toolkit

_AI

_{Cloud & DevOps}

_Product

◆ Signal

snake

◆ Next

+ shipped   production AI features, agents and RAG pipelines
+ shipped   fully automated cloud + mobile release platforms
! building  more autonomous, tool-using agents
! building  LLMOps: evaluation, observability and cost control
? exploring fine-tuning, multimodal models, GitOps at scale

◆ Off the clock

📚 Reading · ⚽ Football · 🚀 Whatever's new in AI this week

Have an idea that needs AI and infrastructure?
athulsreee.com · athult2000@gmail.com

Navigation Menu