100% Open Source • Apache 2.0 • Go-Powered OTel Proxy

Chop the noise before your vendor bills you for it. See what you save, in dollars.

Telemetry Chopper is an open-source OpenTelemetry proxy that drops health checks, debug logs and other noise at the edge, redacts PII inside your perimeter, and prices every dropped byte at your own contract rates.

Quickstart with Docker Run the Sandbox

git clone https://github.com/PRIYAM232/telemetry-chopper.git && cd telemetry-chopper && docker compose -f deploy/docker/docker-compose.yaml up -d --build

Demo data

prod-us-east-1 · 12 collectors

Ingested

18,432.6 GB

Dropped

8,127.4 GB

Saved (MTD)

$14,629.32

Throughput · last 30s

ForwardedChopped

Rule hits

44.1% volume reduction

DROPGET /healthz×1,200
DROPseverity=DEBUG×1,517
REDACTssn → [REDACTED_PATTERN]×1,834
SAMPLEcheckout.span @ 10%×2,151

Built natively for the OpenTelemetry ecosystem

OTLP Compliant

gRPC :4317 · HTTP :4318

Low-Latency Overhead

Zero-alloc regex matching

Self-Hosted Security

Runs inside your cluster

Deterministic Engine

Explicit rules, no ML

Cost calculator

See what your noise is costing you.

Plug in your daily volume and vendor for a quick estimate. It's a simplified version of the dashboard's model: one blended price per GB plus cloud egress. The dashboard prices each signal and indexed events separately, at your own contract rates.

Daily telemetry volume1.0 TB/day

Primary vendor

~$1.80/GB · Blended ingest + 15-day indexing estimate

Estimated noise reduction40%

Estimates use illustrative list rates plus $0.09/GB cloud egress. Configure your exact contract in the Rate Card engine.

Estimated monthly savings

$22,680

Annual savings

$272,160

Cloud egress savings / mo

$1,080

Data chopped per month12 TB

Forwarded to DatadogChopped at the edge

Overage headroom

If your commitment equals today's volume, this reduction leaves room for a +67% traffic spike before overage billing.

Features

A FinOps control plane for your telemetry pipeline.

Everything you need to shrink vendor bills without flying blind during an incident.

Deterministic Rule Engine

No black-box ML magic. You retain 100% explicit control over what gets dropped: nothing disappears unless a rule you wrote says so.

DROPTHROTTLEREDACTSAMPLE

Custom Vendor Rate Cards & Split-Metric Calculations

Configure your exact enterprise contract rates. Model savings across raw ingestion vs. indexed search events.

Logs ingest$0.10 / GB

Indexed log events$1.70 / M

Indexed spans$1.70 / M

Overage & Egress Protection

Cap noisy-tenant spikes with THROTTLE rules and track billable volume against your commitment, so you see overage risk before the invoice does, while cutting outbound cloud network transfer fees.

RawAfter ChopperContract commit

In-Cluster PII Scrubbing

Redact credit card numbers, SSNs, and authorization tokens before they ever leave your Kubernetes cluster.

- ssn=078-05-1120

+ ssn=[REDACTED_PATTERN]

- card=4111 1111 1111 1111

+ card=[REDACTED_PATTERN]

+ authorization=[REDACTED]

How it works

One hop between your apps and your bill.

Telemetry Chopper drops in as a standard OTLP endpoint. The data plane touches your telemetry; the control plane never does.

App / OTel SDKs

Your services emit logs, traces, and metrics over OTLP — no code changes, just point the exporter at Chopper.

LogsTracesMetrics

Telemetry Chopper Proxy

Edge filtering and redaction in a Go-powered collector. Rules sync from the control plane in seconds, with zero restarts.

DROPSAMPLEREDACTTHROTTLE

Datadog, Splunk, New Relic or any OTLP backend

A smaller, cleaner stream reaches your vendor: the signals you keep, without the noise your rules remove.

Noise removedPII redacted

Control Plane (Next.js + Postgres)⇄poll rules every 10s · push match/drop stats · fails open on last-known ruleset

Comparison

Telemetry Chopper vs. alternatives

Collectors move data. Enterprise pipelines charge for it. AI reducers guess. Chopper gives you explicit rules and a dollar figure for every one of them.

Feature comparison of Telemetry Chopper and alternative telemetry pipelines
Feature	Telemetry Chopper	Vector / OTel	Cribl	Grepr	Sawmills
100% Open Source	Yes	Yes	No	NoTools only	NoSome components
Real-Time Dollar Dashboard	Yes	No	PartialAdd-on pack	No	PartialMonthly reports
Explicit Deterministic Rules	Yes	Yes	Yes	Partial	PartialAI-recommended
Custom Rate Card Engine	Yes	No	No	No	No
Free / Self-Hosted	Yes	Yes	PartialFree ≤1 TB/day	PartialSaaS-first	NoFrom $2,500/mo

YesSupported out of the box
PartialLimited, add-on, or with conditions (see note)
NoNot offered or not publicly documented

Based on public vendor documentation as of September 2026. View sources

Vector / OTel

Cribl

Grepr

Sawmills

Product names are trademarks of their respective owners and are used for identification only. Capabilities change: verify with each vendor before making a purchasing decision. Spot an error? Open an issue.

Developer quickstart

From zero to chopping in five minutes.

Run it anywhere you run containers. Create rules in the dashboard, and every collector picks them up within seconds, with no restarts.

Chopper receives telemetry over OTLP: point your OpenTelemetry SDKs or Collector at it. Data sent through a vendor's own agent goes straight to the vendor and bypasses it.

terminal

1$ git clone https://github.com/PRIYAM232/telemetry-chopper.git
2$ cd telemetry-chopper
3
4# Postgres → migrate + seed → Control Plane (:3000) → Collector (:4317/:4318)
5$ docker compose -f deploy/docker/docker-compose.yaml up -d --build
6
7# Dashboard: http://localhost:3000

GET /api/v1/policies/{fleetId} synced every 10s

Rules you create in the dashboard, as every collector in the fleet receives them (abridged).

1{
2  "fleet_id": "3f0c…",
3  "rules": [
4    {
5      "name": "drop-health-checks",
6      "isActive": true,
7      "actionType": "DROP",
8      "targetSignal": "TRACES",
9      "conditionField": "http.route",
10      "conditionOp": "REGEX_MATCH",
11      "conditionValue": "^/(health|healthz|ready)$"
12    },
13    {
14      "name": "redact-ssn",
15      "isActive": true,
16      "actionType": "REDACT",
17      "targetSignal": "LOGS",
18      "conditionField": "log.body",
19      "conditionOp": "REGEX_MATCH",
20      "conditionValue": "\\b\\d{3}-\\d{2}-\\d{4}\\b"
21    },
22    {
23      "name": "drop-debug-logs",
24      "isActive": true,
25      "actionType": "DROP",
26      "targetSignal": "LOGS",
27      "conditionField": "log.severity",
28      "conditionOp": "EQUALS",
29      "conditionValue": "DEBUG"
30    }
31  ]
32}