Procedure

Mobile builds sound simple. Until CI servers, signing, and infra enter the picture. Expo EAS removes the entire CI layer — not just optimizes it. Here's how 👇 procedure.tech/blogs/automate…

PostgreSQL didn't slow down because of bad queries. It slowed down because 92% of a 123GB table belonged to one tenant. LIST + HASH partitioning fixed it: 1. updates 30–60s → 5–7s 2. lookups ~10× faster 3. no query changes 👉 Read: procedure.tech/blogs/database…

We broke down how to do this safely in Part 2 of our EAS series. Mastering Expo EAS: Submit, OTA Updates, and Workflow Automation -procedure.tech/blogs/masterin…

The trick is discipline: - channels - runtimeVersion - env-aware workflows Without those, OTA becomes chaos.

EAS Submit removes the store friction. OTA updates remove the "small fix = big delay" problem.

Build automation helps. But teams still get stuck on: - manual store uploads - slow hotfixes - release-day anxiety

If your mobile release still feels like an event, your pipeline isn't done yet 🧵

We’re hiring a Forward Deployed Engineer📢 Work directly with clients, own problems end-to-end, and ship what actually moves the needle. Remote | India Apply now: procedure.keka.com/careers/jobdet…

Read how LIST + HASH partitioning actually fixed a 123GB bottleneck: procedure.tech/blogs/database…

Most Django apps still log audits into DB tables. But is that really where they belong? What do you think? Where should audit trails live in a Django app? 👉 Read the full blog: procedure.tech/blogs/structur…

We’re hiring an Office Manager at our Mumbai Office (Bhandup W), who can run ops, manage stakeholders & keep things moving. If you’re built for execution, this role is yours. Apply now: procedure.keka.com/careers/jobdet…

Scalable frontend architecture is not about React. It's about making the right decisions before the repo becomes hostile. Full step-by-step guide 👇 procedure.tech/blogs/blog-sca…

Full fundamentals guide here 👇 procedure.tech/blogs/observab…

OpenTelemetry helps standardize this data so you're not locked into one vendor.

Alerts and dashboards only work after these signals are right.

Traces → follow a request across services Logs → capture what actually happened Metrics → measure impact over time

Most production incidents aren't mysterious — we just don't know where to look 🧵

That's the gap observability fills.

🚨We’re hiring AI-native full-stack engineers who can ship across the stack. If you’re building with LLMs, RAG, and modern stacks - we want you on this team. Location: Remote Experience: 2 - 5 yrs Apply now: procedure.keka.com/careers/jobdet…

Full breakdown + code examples here 👇procedure.tech/blogs/how-to-p…

When to use Workers: 1. Data-heavy dashboards 2. Image/video processing 3. Client-side ML

JavaScript UI freezes are rarely "framework issues". They're architectural mistakes 🧵