On-Call Health

Product Introduction

1. Definition: On-Call Health is an open-source DevOps/SRE wellness tool that monitors engineer burnout risk by aggregating objective workload data (incidents, tickets, code commits) and subjective sentiment signals. It falls under the IT operations analytics and developer productivity categories.
2. Core Value Proposition: It proactively identifies unsustainable on-call workloads using personalized risk scoring to prevent engineer burnout, reduce attrition, and maintain team productivity through data-driven well-being interventions.

Main Features

1. Multi-Source Signal Integration:

   • How it works: Connects via APIs to ingest real-time data from Rootly/PagerDuty (incidents), Linear/Jira (ticket workload), GitHub (after-hours activity), and Slack (communication patterns). Uses OAuth 2.0 for authentication.
   • Technology: Apache 2.0-licensed Python/Node.js backend with configurable data pipelines for ETL (Extract, Transform, Load).

2. Automated Sentiment Surveys:

   • How it works: Deploys brief, anonymized Slack surveys at configurable intervals (e.g., post-incident) to capture self-reported stress levels. Responses are weighted against historical baselines.
   • Technology: Slack Bolt API integration with time-series databases (e.g., PostgreSQL) for longitudinal sentiment tracking.

3. Dynamic Risk Scoring Engine:

   • How it works: Generates 0–100 risk scores per engineer using machine learning models that correlate workload volume, after-hours work frequency, incident stress, and survey data. Scores categorize risk tiers (e.g., 50–74 = "Early Intervention").
   • Technology: Regression-based anomaly detection comparing current workloads against 90-day personal/team baselines.

Problems Solved

1. Pain Point: Prevents "silent burnout" where engineers hide unsustainable workloads until quitting, causing talent loss and incident response gaps.
2. Target Audience:
   • SRE/DevOps leads managing on-call rotations
   • Engineering managers at scaling tech companies
   • Incident commanders in high-velocity SaaS teams
3. Use Cases:
   • Detecting chronic overload in senior engineers handling complex incidents
   • Validating workload distribution during team scaling
   • Auditing on-call fairness post-merger/acquisition

Unique Advantages

1. Differentiation: Unlike PagerDuty Analytics (limited to incident metrics) or generic wellness tools (e.g., Calm), On-Call Health combines technical telemetry with human context for burnout prediction.
2. Key Innovation: Personalized baselines eliminate one-size-fits-all thresholds—risk scores reflect individual capacity changes, not peer comparisons.

Frequently Asked Questions (FAQ)

1. How does On-Call Health calculate burnout risk?
   It combines incident frequency, after-hours work (GitHub timestamps), ticket volume, sentiment surveys, and communication patterns into a weighted algorithm benchmarked against personal historical data.

2. Is On-Call Health compliant with data privacy regulations?
   Yes, it adheres to GDPR/CCPA via anonymized data processing, optional PII scrubbing, and strict access controls for sentiment data.

3. Can it integrate with our existing Jira Cloud setup?
   Absolutely—it supports OAuth 2.0 integrations with Jira Cloud API, syncing ticket assignments, resolution times, and priority levels.

4. What makes this better than manual wellness checks?
   It removes stigma by automating data collection, provides objective benchmarks (vs. subjective guesses), and flags issues 4–6 weeks earlier via trend analysis.

5. How quickly can we deploy On-Call Health?
   As open-source software, it deploys in <2 hours via Docker/Kubernetes, with pre-built connectors for PagerDuty, Slack, and GitHub.