Engineering blog

From the factory floor
to the codebase.

Real engineering articles from the team building edge MLOps infrastructure. No fluff. No rephrased documentation. Just what we learned shipping models to ARM devices.

All articles
Engineering July 15, 2025 · 12 min read

Why your MLOps pipeline fails silently at the edge

Three weeks, six cameras, zero alerts. The architecture pattern that prevents it.

RG
Raghunathareddy GR
CEO & Founder
Read article →
Engineering June 28, 2025 · 9 min read

Building an offline-first sync protocol for edge ML agents

How we designed the delta-sync protocol that stays consistent across 2G connections and 6-hour outages.

RG
Raghunathareddy GR
CEO & Founder
Read article →
Research June 14, 2025 · 11 min read

Statistical drift detection without the cloud: KL divergence at the edge

Running Kullback-Leibler divergence checks on a Jetson Nano. Spoiler: it works, if you batch it right.

RG
Raghunathareddy GR
CEO & Founder
Read article →
DevOps May 30, 2025 · 7 min read

Canary deployments to heterogeneous ARM fleets

Not all Jetson Nanos are equal. How we target rollouts by hardware capability, not just device count.

RG
Raghunathareddy GR
CEO & Founder
Read article →
Engineering May 12, 2025 · 8 min read

How we shrunk the mlops-agent binary to under 8MB

Go, musl, UPX, and a lot of feature flags. A forensic walkthrough of every KB stripped from the build.

RG
Raghunathareddy GR
CEO & Founder
Read article →
Research April 22, 2025 · 14 min read

Latency benchmarks: ONNX vs TFLite vs TensorRT on Jetson Orin

Real numbers from real hardware, at sustained load, with thermal throttling data. Not cloud VMs.

RG
Raghunathareddy GR
CEO & Founder
Read article →

Get new articles in your inbox

One email per article. No marketing. Unsubscribe anytime.

© 2025 MLOps.dev Technologies Pvt. Ltd. · Raghunathareddy GR · www.mlops.dev
Discord