Developer Articles | TechForDev

Latest AI / ML JavaScript Python React Next.js Web Dev DevOps Cloud

p95 latency was 1.68 seconds. I didn't know until I built trace-stats.

Mukunda Rao Katta3d ago • 3 min read

p95 latency was 1.68 seconds. I didn't know until I built trace-stats.

A zero-dep Python lib that computes p50/p90/p95/p99 percentile statistics on any numeric field in an...

#devchallenge#hermesagentchallenge#agents#observability

3 0

The two-line Hermes agent logger I wish existed a month ago

Mukunda Rao Katta3d ago • 3 min read

The two-line Hermes agent logger I wish existed a month ago

A zero-dep Python context manager that writes one JSONL line per agent step with start time, duratio...

#devchallenge#hermesagentchallenge#agents#observability

3 0

My Hermes agent spent how much? trace-cost for JSONL audit logs

Mukunda Rao Katta3d ago • 3 min read

My Hermes agent spent how much? trace-cost for JSONL audit logs

A zero-dep Python lib that reads any agent JSONL trace and computes cost totals by model, lane, and ...

#devchallenge#hermesagentchallenge#agents#observability

3 1

I ran three Hermes agents in parallel. trace-merge stitched them into one timeline I could read.

Mukunda Rao Katta3d ago • 5 min read

I ran three Hermes agents in parallel. trace-merge stitched them into one timeline I could read.

A zero-dep Python lib that takes N agent JSONL audit logs and emits one chronologically merged strea...

#devchallenge#hermesagentchallenge#agents#observability

3 0

Pipe your Hermes Agent audit log into Datadog in 5 lines (no OTel SDK required)

Mukunda Rao Katta3d ago • 5 min read

Pipe your Hermes Agent audit log into Datadog in 5 lines (no OTel SDK required)

A zero-dep Python lib that turns agent JSONL audit logs into OTLP/JSON spans you can POST to any Ope...

#devchallenge#hermesagentchallenge#observability#agents

3 0

I had 800 lines of Hermes agent audit log. trace-tree turned it into a tree I could read.

Mukunda Rao Katta3d ago • 5 min read

I had 800 lines of Hermes agent audit log. trace-tree turned it into a tree I could read.

A zero-dep Python lib that reads agent JSONL audit logs and prints a real ASCII tree of what the age...

#devchallenge#hermesagentchallenge#agents#observability

4 0

Sidemark: Active Telemetry Comments for C#

David Whitney5h ago • 4 min read

Sidemark: Active Telemetry Comments for C#

OpenTelemetry has quietly become table stakes. That's a good thing, but if you've instrumented a rea...

#dotnet#opensource#opentelemetry#observability

1 0

One Hermes tool call took 4.8 seconds. The others averaged 900ms. trace-anomaly found it.

Mukunda Rao Katta3d ago • 3 min read

One Hermes tool call took 4.8 seconds. The others averaged 900ms. trace-anomaly found it.

A zero-dep Python lib that uses IQR-based outlier detection to find slow steps, expensive calls, and...

#devchallenge#hermesagentchallenge#agents#observability

3 0

5000 events, one worker, one bug: trace-filter for agent JSONL traces

Mukunda Rao Katta3d ago • 4 min read

5000 events, one worker, one bug: trace-filter for agent JSONL traces

A zero-dep Python lib that filters events from agent JSONL logs by lane, kind, time range, or any fi...

#devchallenge#hermesagentchallenge#agents#observability

0 0

I Got Tired of Wiring Grafana + Loki + Tempo — So I Built a One-Command Observability Stack

Sulthon Zainul Habib9h ago • 4 min read

I Got Tired of Wiring Grafana + Loki + Tempo — So I Built a One-Command Observability Stack

Setting up the three pillars of observability should not take a weekend. Here is how I made it take ...

#observability#devops#opentelemetry#docker

0 0

I logged 300 Hermes runs to one file. trace-session-split cut it into 300.

Mukunda Rao Katta3d ago • 2 min read

I logged 300 Hermes runs to one file. trace-session-split cut it into 300.

A zero-dep Python lib that splits a multi-session JSONL log by run_id, session_id, or any field into...

#devchallenge#hermesagentchallenge#agents#observability

0 0

How sync.Pool Helped Me Stabilize p99 Latency in a High-Throughput Log Processing Pipeline

amir8h ago • 12 min read

How sync.Pool Helped Me Stabilize p99 Latency in a High-Throughput Log Processing Pipeline

A real-world story about GC pressure, JSON processing, network buffers, and how object pooling impro...

#go#performance#backend#observability

0 0

Two states weren't enough. Here's why I added WATCH.

Lazypl822d ago • 6 min read

Two states weren't enough. Here's why I added WATCH.

A teammate had two browser tabs open and one Slack thread waiting on him. The checkout API had just...

#devops#sre#deployment#observability

0 0

What GitHub Uses eBPF For (and the Layer They Have Not Ported Yet)

Ingero Team4d ago • 5 min read

What GitHub Uses eBPF For (and the Layer They Have Not Ported Yet)

Three eBPF patterns hyperscalers run in production today, mapped to the equivalent patterns on the.....

#devops#linux#opensource#observability

0 0

I Built chanprobe Because My Go Queues Were Invisible

Pavel Sanikovich6d ago • 7 min read

I Built chanprobe Because My Go Queues Were Invisible

A small story about Go channels, hidden backpressure, and why I built an observable bounded queue fo...

#go#observability#opensource

2 0

Distributed tracing across FastAPI and Celery with OpenTelemetry the part nobody shows you

Jayanth5d ago • 2 min read

Distributed tracing across FastAPI and Celery with OpenTelemetry the part nobody shows you

Every OpenTelemetry tutorial ends at "add the middleware to FastAPI." That gets you traces for HTTP....

#python#observability#distributedsystems#backend

0 0

From Kernel Scheduler to Python Source Line: Tracing a GPU Stall End to End

Ingero Team5h ago • 6 min read

From Kernel Scheduler to Python Source Line: Tracing a GPU Stall End to End

TL;DR A GPU that reports 97% utilization can still be the slowest part of a training step,...

#ebpf#gpu#python#observability

0 0

Translating LLM Telemetry Between OpenInference and OTel GenAI with Rust

Mukunda Rao Katta3d ago • 5 min read

Translating LLM Telemetry Between OpenInference and OTel GenAI with Rust

otel-genai-bridge-rs translates LLM span attribute names between OpenInference semconv and OTel GenA...

#hermeschallenge#ai#rust#observability

0 0

Manas Sharma1d ago • 10 min read

How to Monitor AI Agents in Production

How to make AI agents observable in production: OTel instrumentation, telemetry backends, cost track...

#aiagents#opentelemetry#observability#llm

1 3

I Turned on Agent Tracing for 30 Days. 4 Hidden Bottlenecks Were Eating 47% of My Tokens.

Ken Imoto1d ago • 7 min read

I Turned on Agent Tracing for 30 Days. 4 Hidden Bottlenecks Were Eating 47% of My Tokens.

A production Claude agent had been quietly burning 5.2M tokens a month. I turned on per-call tracing...

#ai#agents#observability#performance

0 0

Chronos vs Toto: Zero-Shot Forecasting Benchmark Results

Parseable Team2d ago • 14 min read

Chronos vs Toto: Zero-Shot Forecasting Benchmark Results

Introduction Good forecasts help with capacity planning and quieter alerts. But one...

#chronos#dataengineering#observability

1 0

OpenTelemetry Is Now a CNCF Graduate — and It's Coming for Your AI Stack

Andrew Kew6d ago • 3 min read

OpenTelemetry Is Now a CNCF Graduate — and It's Coming for Your AI Stack

OpenTelemetry graduated as a CNCF project on May 21, 2026. That's not just a badge — it's the formal...

#opentelemetry#observability#devops#ai

1 0

Unlocking Insights with Observability: My Journey with OpenTelemetry

Naveen Malothu1d ago • 2 min read

Unlocking Insights with Observability: My Journey with OpenTelemetry

Learn how to use OpenTelemetry to gain insights into your applications and services, and improve obs...

#observability#devops#monitoring

1 1

AI SRE and AI DevOps: different problems, one reliability stack

Divyansh 2d ago • 6 min read

AI SRE and AI DevOps: different problems, one reliability stack

Vendors and headlines often blur "AI for operations" into one bucket. In practice, two distinct...

#sre#devops#platformengineering#observability

2 0

Tech Articles

p95 latency was 1.68 seconds. I didn't know until I built trace-stats.

The two-line Hermes agent logger I wish existed a month ago

My Hermes agent spent how much? trace-cost for JSONL audit logs

I ran three Hermes agents in parallel. trace-merge stitched them into one timeline I could read.

Pipe your Hermes Agent audit log into Datadog in 5 lines (no OTel SDK required)

I had 800 lines of Hermes agent audit log. trace-tree turned it into a tree I could read.

Sidemark: Active Telemetry Comments for C#

One Hermes tool call took 4.8 seconds. The others averaged 900ms. trace-anomaly found it.

5000 events, one worker, one bug: trace-filter for agent JSONL traces

I Got Tired of Wiring Grafana + Loki + Tempo — So I Built a One-Command Observability Stack

I logged 300 Hermes runs to one file. trace-session-split cut it into 300.

How sync.Pool Helped Me Stabilize p99 Latency in a High-Throughput Log Processing Pipeline

Two states weren't enough. Here's why I added WATCH.

What GitHub Uses eBPF For (and the Layer They Have Not Ported Yet)

I Built chanprobe Because My Go Queues Were Invisible

Distributed tracing across FastAPI and Celery with OpenTelemetry the part nobody shows you

From Kernel Scheduler to Python Source Line: Tracing a GPU Stall End to End

Translating LLM Telemetry Between OpenInference and OTel GenAI with Rust

How to Monitor AI Agents in Production

I Turned on Agent Tracing for 30 Days. 4 Hidden Bottlenecks Were Eating 47% of My Tokens.

Chronos vs Toto: Zero-Shot Forecasting Benchmark Results

OpenTelemetry Is Now a CNCF Graduate — and It's Coming for Your AI Stack

Unlocking Insights with Observability: My Journey with OpenTelemetry

AI SRE and AI DevOps: different problems, one reliability stack