---
title: "Sifflet"
description: "Faster monitoring execution, with an AI assistant to help investigate incidents."
canonical_url: "https://www.rohanchaudhari.fr/portfolio/sifflet"
md_url: "https://www.rohanchaudhari.fr/portfolio/sifflet.md"
last_updated: "2026-09-22"
---

# Sifflet

Faster monitoring execution, with an AI assistant to help investigate incidents.

## Context

Production experience · 2025–2026

[Sifflet website](https://www.siffletdata.com/)

Software engineering on the monitoring team, September 2025 to February 2026.

## My role

Software Engineer — Monitoring Team

Orchestration improvements and an AI-driven incident-analysis assistant within the wider monitoring product.

## Problem

Monitoring has to produce useful signals promptly. When an incident needs investigation, logs and metadata also need to become context an engineer can act on.

## Result and evidence

55% reduction in end-to-end monitoring execution latency through orchestration improvements.

## Boundaries

The latency result applies to the orchestration work. No separate accuracy, adoption, or incident-resolution metric is claimed for the assistant. The diagram is a simplified illustration of the work areas.

[Read the illustrated case study](/portfolio/sifflet).

## Sitemap

See the full [Markdown sitemap](/sitemap.md), [LLM index](/llms.txt), or [contact page](/contact.md).