Site Reliability Engineer with 4+ years operating mission-critical systems, currently at Banco Falabella (SRE automation & scalability cell). I run services on Kubernetes with Helm and Argo CD, automate operations in Python and Ansible, manage multi-cloud infrastructure with Terraform, and work incidents from alert to postmortem. Before that, three years running observability for Fortune 500 clients. I also publish open-source Python & MCP tools. I like to keep the lights on.
apiVersion: reliability.fabian.dev/v1
kind: SiteReliabilityEngineer
metadata:
name: fabian-munoz
namespace: santiago-chile
labels:
role: sre
experience: 4y+
availability: remote-and-international-travel
spec:
currentRole: Site Reliability Engineer @ Banco Falabella (Oct 2025 – present)
previousRole: Site Reliability Engineer @ Innfinit SpA — Fortune 500 Consultancy (Nov 2022 – Oct 2025)
education:
- Computer Engineering (Ingeniería en Informática), Duoc UC · 2025
- Programming Analyst, Duoc UC · 2024
languages:
- Spanish (native)
- English (professional working proficiency)
focus:
- GitOps & Kubernetes Operations (Helm, Argo CD, Vault)
- High-Availability Observability & SLI/SLO Engineering
- AI-Assisted Incident Response & Approval-Gated Automation
- Multi-Cloud Infrastructure as Code (Terraform)
status:
phase: Running
motto: "I like to keep the lights on."flowchart LR
A["📦 GitOps Delivery<br/>Helm · Argo CD · Vault · WAF"] --> B["☸️ Production Workloads<br/>K8s · Kafka · Multi-Cloud"]
B --> C["📡 Telemetry & Signals<br/>Datadog · Prometheus · Grafana · Loki"]
C --> D{"🎯 SLO Breached?"}
D -- no --> E["😴 Sleep Well"]
D -- yes --> F["🤖 AI Incident Agent<br/>Correlates alerts & deploys"]
F --> G["🔐 Approval-Gated Ops<br/>Rundeck · Ansible"]
G --> H["📝 Postmortem & RCA"]
H -. "harden & automate" .-> A
classDef core fill:#312E81,stroke:#818CF8,color:#F8FAFC
classDef ok fill:#064E3B,stroke:#10B981,color:#ECFDF5
classDef alert fill:#7F1D1D,stroke:#F87171,color:#FEF2F2
classDef ai fill:#4C1D95,stroke:#C084FC,color:#FAF5FF
class A,B,C core
class E ok
class F,G ai
class H alert
Co-building an AI-assisted incident platform where an LLM agent correlates alerts and deployments across Datadog, Loki, and cloud, proposing approval-gated Rundeck remediation jobs. Also built a multi-agent issue-to-Argo CD release pipeline piloted on self-hosted open-weight models.
Operating Kubernetes services and automated ETL workloads under GitOps with Helm and Argo CD. Took an internal Python MCP server from proof of concept to a hardened production service: build pipeline, Helm chart, Argo CD, Vault secrets, and Cloudflare WAF.
Operating observability for Kafka event streaming in high-availability transaction processing, cross-site disaster recovery planning, migrating Rundeck automations to Ansible Automation Platform, and managing Datadog, GCP, Azure, and Cloudflare with Terraform.
Three years delivering reliability and observability across Liberty Mutual US (RCA, postmortems, Splunk → Datadog migration), Sovos (observability lead — Prometheus Operator & Grafana Cloud across dev/QA/prod), Red Salud (HA Prometheus/Grafana/Loki + Slack/Teams/PagerDuty), and Liberty Andes (3-country Datadog-as-Code with Terraform).
| Project | Description | Links |
|---|---|---|
| 🧠 madeleine | MCP server in Python that recalls related memories for AI agents from context such as place or time. Published on PyPI & the Official MCP Registry. | |
| 🎙️ Voice Coach MCP | Local Python MCP server for pronunciation and speaking practice with on-device audio analysis. Published on PyPI. | |
| 🎵 Setlist to YouTube Music | FastAPI & React app with OAuth, background jobs, Docker, GitHub Actions CI, and 240+ tests. |
| Domain | Stack |
|---|---|
| Languages & Automation | Python · Bash · Ansible Automation Platform · Rundeck · GitHub Actions · GitLab CI |
| Platforms & GitOps | Linux · Kubernetes · Helm · Argo CD · Argo Workflows · Docker · HashiCorp Vault |
| Cloud & IaC | AWS · Azure · GCP · Terraform · CloudFormation · Cloudflare |
| Observability & Incidents | Datadog · Prometheus · Grafana · Loki · CloudWatch · Langfuse · SLI/SLO · Postmortems |
| Data & Messaging | Apache Kafka (Confluent) · PostgreSQL |
🐍 A snake eating my contributions — the only acceptable kind of data loss.
🚨 Need an SRE? Click to open the runbook
# Step 1 — Identify the issue
$ echo "Need reliable K8s, clean observability, or AI-assisted ops"
# Step 2 — Page the engineer
$ page --to "Fabián Muñoz" \
--severity "let's-talk" \
--channel https://linkedin.com/in/fabianimv \
--portfolio https://fabianimv.github.io/portfolio \
--fallback fabianignaciomv@gmail.com
# Step 3 — Resolution
✔ GitOps & IaC automated
✔ SLOs defined & alert noise cut
✔ Incident response accelerated
✔ Lights stay on




