Hi! I'm

Federico Calò

Software Developer | Technical Writer

I create modern web applications and custom digital tools to help businesses grow through technological innovation. My passion is combining computer science and economics to generate real value.

Contact Me

About Me

My passion for computer science was born at the Technical Commercial Institute of Maglie, where I discovered the power of programming and the fascination of creating digital solutions. From the start, I understood that computer science was not just code, but an extraordinary tool for turning ideas into reality.

During my studies in Business Information Systems, I began to interweave computer science and economics, understanding how technology can be the engine of growth for any business. This vision accompanied me to the University of Bari, where I obtained my degree in Computer Science, deepening my technical skills and passion for software development.

Today I put this experience at the service of businesses, professionals and startups, creating tailor-made digital solutions that automate processes, optimize resources and open new business opportunities. Because true innovation begins when technology meets the real needs of people.

My Skills

Data Analysis & Predictive Models

I transform data into strategic insights with in-depth analysis and predictive models for informed decisions

Process Automation

I create custom tools that automate repetitive operations and free up time for value-added activities

Custom Systems

I develop tailor-made software systems, from platform integrations to customized dashboards

const federico = {
  nome: "Federico Calò",
  ruolo: "Sviluppatore Software",
  città: "Bari, Italia",
  missione: "Aiutare attraverso l'informatica",
  passioni: [
    "Codice Pulito",
    "Innovazione",
    "Crescita Continua"
  ]
};

La Mia Missione

Credo fermamente che l'informatica sia lo strumento più potente per trasformare le idee in realtà e migliorare la vita delle persone.

🚀

Democratizzare la Tecnologia

La mia missione è rendere l'informatica accessibile a tutti: dalle piccole imprese locali alle startup innovative, fino ai professionisti che vogliono digitalizzare la propria attività. Ogni realtà merita di sfruttare le potenzialità del digitale.

💡

Unire Informatica ed Economia

Non è solo questione di scrivere codice: è capire come la tecnologia possa generare valore reale. Intrecciando competenze informatiche e visione economica, aiuto le attività a crescere, ottimizzare processi e raggiungere nuovi traguardi di efficienza e redditività.

🎯

Creare Soluzioni su Misura

Ogni attività è unica, e così devono esserlo le soluzioni. Sviluppo strumenti personalizzati che rispondono alle esigenze specifiche di ciascun cliente, automatizzando processi ripetitivi e liberando tempo per ciò che conta davvero: far crescere il business.

Trasforma la Tua Attività con la Tecnologia

December 2024

View

Master SQL

RoadMap.sh

Novembre 2024

View

Oracle Certified Foundations Associate

Oracle

October 2024

View

People Leadership Credential

Connect

Settembre 2024

💻 Languages & Technologies

☕Java

🐍Python

📜JavaScript

🅰️Angular

⚛️React

🔷TypeScript

🗄️SQL

🐘PHP

🎨CSS/SCSS

🔧Node.js

🐳Docker

🌿Git

💼

12/2024 - Presente

Custom Software Engineering Analyst

Accenture

Bari, Puglia, Italia · Ibrida Analisi e sviluppo di sistemi informatici attraverso l'utilizzo di Java e Quarkus in Health and Public Sector. Formazione continua su tecnologie moderne per la creazione di soluzioni software personalizzate ed efficienti e sugli agenti.

💼

06/2022 - 12/2024

Analista software e Back End Developer Associate Consultant

Links Management and Technology SpA

Esperienza nell'analisi di sistemi software as-is e flussi ETL utilizzando PowerCenter. Formazione completata su Spring Boot per lo sviluppo di applicazioni backend moderne e scalabili. Sviluppatore Backend specializzato in Spring Boot, con esperienza in progettazione di database, analisi, sviluppo e testing dei task assegnati.

💼

02/2021 - 10/2021

Programmatore software

Adesso.it (prima era WebScience srl)

Esperienza nell'analisi AS-IS e TO-BE, evoluzioni SEO ed evoluzioni website per migliorare le performance e l'engagement degli utenti.

🎓

2018 - 2025

Laurea in Informatica

Università degli Studi di Bari Aldo Moro

Bachelor's degree in Computer Science, focusing on software engineering, algorithms, and modern development practices.

📚

2013 - 2018

Diploma - Sistemi Informativi Aziendali

Istituto Tecnico Commerciale di Maglie

Technical diploma specializing in Business Information Systems, combining IT knowledge with business management.

Contattami

Hai un progetto in mente? Parliamone! Compila il form qui sotto e ti risponderò al più presto.

* Campi obbligatori. I tuoi dati saranno utilizzati solo per rispondere alla tua richiesta.

FinOps for Kubernetes: Rightsizing, Spot Instances and Cost Reduction

68% of organizations using Kubernetes in production spend 20-40% more than necessary (CNCF Cost Survey 2026). Not because Kubernetes is inefficient, but because generous and easy provisioning, and rightsizing requires operational discipline. The result and cluster with average CPU usage at 15-20% and memory at 30-40%: resources paid but don't use.

FinOps for Kubernetes and the practice of measuring, optimizing and governing cluster spending. In this article we will see how to use Kubecost for the visibility of costs for namespaces, such as the VPA helps in rightsizing automatic requests, how Karpenter optimizes node bin-packing with spot instances, and which policies to implement at the organizational level to maintain costs under control over time.

What You Will Learn

Install and configure Kubecost for namespace/team cost visibility
VPA in Off mode: Automatic rightsizing recommendations without disruption
Karpenter: Consolidation for optimal bin-packing of nodes
Spot instances on AWS/GCP/Azure with Karpenter and interrupt management
ResourceQuota as a team budgeting tool
Alert Prometheus for waste and cost anomalies
How to build a chargeback system for teams
Benchmark: real savings achievable with each technique

Why Kubernetes Tends to be Overprovisioned

Before optimizing, you need to understand the causes of the problem:

Conservative Requests: Teams set requests high "to be safe" (e.g. 1 CPU for an app that uses 0.1). Requests determine the number of nodes needed
Limits = requests: Many CI/CD templates set limits equal to requests, creating systematic over-provisioning
Empty namespaces: Development namespace with "Hello World" Pods that stay on 24/7
Oversized knots: Large cloud instances with inefficient bin-packing (a 2 vCPU Pod on a 32 vCPU node)
No visibility for teams: If teams don't see their cost, they have no incentive to optimize

Kubecost: Visibility of Namespace Costs

Kubecost It is the most used FinOps tool in the Kubernetes ecosystem. Collects metrics from Prometheus, retrieves cloud instance pricing (AWS, GCP, Azure) and calculate the cost allocated for namespace, deployment, service account and label.

Kubecost installation

# Installa Kubecost con Helm
helm repo add kubecost https://kubecost.github.io/cost-analyzer/
helm repo update

helm install kubecost kubecost/cost-analyzer \
  --namespace kubecost \
  --create-namespace \
  --set kubecostToken="your-token-here" \
  --set prometheus.enabled=true \
  --set grafana.enabled=true \
  --set global.prometheus.enabled=false \
  --set global.prometheus.fqdn="http://prometheus.monitoring.svc:9090"

# Port-forward per accedere alla UI
kubectl port-forward svc/kubecost-cost-analyzer 9090:9090 -n kubecost &
# Apri: http://localhost:9090

# Alternativa open-source senza token: OpenCost
helm install opencost opencost/opencost \
  --namespace opencost \
  --create-namespace

Configure Chargeback for Teams

# Kubecost usa i label Kubernetes per il cost allocation
# Configura label standard per tutti i workload:

# Nel values.yaml di ogni applicazione (Helm):
podLabels:
  team: "team-alpha"
  cost-center: "CC-2024-ENG"
  environment: "production"
  product: "checkout-service"

# Kubecost mostra automaticamente i costi aggregati per questi label.
# Es: costo totale team-alpha nel mese = CPU + Memoria + Storage + GPU + Networking

# API Kubecost per report automatici (da inviare via email settimanale):
curl "http://kubecost:9090/model/allocation?window=7d&aggregate=label:team&accumulate=true" \
  | jq '.data[0] | to_entries[] | {team: .key, cost: .value.totalCost}'

VPA for Automatic Rightsizing

Il Vertical Pod Autoscaler (VPA) analyzes historical CPU consumption and Pod memory and calculates optimal requests. In mode Off (recommendation without automatic application) is the safest tool for rightsizing: it provides precise numbers without risk of disruption.

# Installa VPA
git clone https://github.com/kubernetes/autoscaler.git
cd autoscaler/vertical-pod-autoscaler
./hack/vpa-up.sh

# Oppure con Helm
helm repo add fairwinds-stable https://charts.fairwinds.com/stable
helm install vpa fairwinds-stable/vpa \
  --namespace vpa \
  --create-namespace

---
# vpa-recommendation.yaml - modalita Off: solo raccomandazioni
apiVersion: autoscaling.k8s.io/v1
kind: VerticalPodAutoscaler
metadata:
  name: api-service-vpa
  namespace: team-alpha
spec:
  targetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: api-service
  updatePolicy:
    updateMode: "Off"   # non applica automaticamente, solo raccomanda
  resourcePolicy:
    containerPolicies:
      - containerName: api
        minAllowed:
          cpu: 50m
          memory: 64Mi
        maxAllowed:
          cpu: "4"
          memory: 8Gi

# Leggi le raccomandazioni dopo 24-48 ore:
kubectl describe vpa api-service-vpa -n team-alpha
# Output:
# Container Recommendations:
#   Container Name:  api
#   Lower Bound:
#     Cpu:     120m
#     Memory:  256Mi
#   Target:               <-- usa questi valori nelle requests
#     Cpu:     250m
#     Memory:  512Mi
#   Upper Bound:
#     Cpu:     800m
#     Memory:  1500Mi
#   Uncapped Target:
#     Cpu:     230m
#     Memory:  480Mi

Script for Rightsizing Analysis in Bulk

# Trova tutti i Deployment con requests > 2x il consumo reale
kubectl get vpa -A -o json | jq -r '
  .items[] |
  .metadata.namespace + "/" + .metadata.name + ": " +
  (.status.recommendation.containerRecommendations[]? |
   "target CPU=" + .target.cpu + " Mem=" + .target.memory)
'

# Script bash per generare report di rightsizing
#!/bin/bash
echo "=== Rightsizing Recommendations ==="
for ns in $(kubectl get ns -o jsonpath='{.items[*].metadata.name}'); do
  echo "--- Namespace: $ns ---"
  kubectl get vpa -n "$ns" -o custom-columns=\
"NAME:.metadata.name,\
TARGET-CPU:.status.recommendation.containerRecommendations[0].target.cpu,\
TARGET-MEM:.status.recommendation.containerRecommendations[0].target.memory" \
    2>/dev/null
done

Karpenter: Consolidation and Spot Instances

Karpenter optimizes costs from two directions: use spot instances for reduce the cost per node by 60-70%, e consolidate the nodes to delete those underutilized (bin-packing).

NodePool with Spot + Consolidation

# karpenter-cost-optimized.yaml
apiVersion: karpenter.sh/v1
kind: NodePool
metadata:
  name: cost-optimized
spec:
  template:
    spec:
      nodeClassRef:
        apiVersion: karpenter.k8s.aws/v1
        kind: EC2NodeClass
        name: default
      requirements:
        # Tipi di istanza ottimali per bin-packing (diverse size)
        - key: node.kubernetes.io/instance-type
          operator: In
          values:
            - m7i.xlarge    # 4 vCPU, 16GB  ~$0.20/h on-demand, ~$0.05/h spot
            - m7i.2xlarge   # 8 vCPU, 32GB
            - m7i.4xlarge   # 16 vCPU, 64GB
            - m7i.8xlarge   # 32 vCPU, 128GB
            - c7i.2xlarge   # CPU-optimized per workload compute-intensive
            - r7i.2xlarge   # Memory-optimized per carichi con alta RAM
        # Mix spot e on-demand con peso
        - key: karpenter.sh/capacity-type
          operator: In
          values:
            - spot       # priorita spot (economico)
            - on-demand  # fallback on-demand
  disruption:
    consolidationPolicy: WhenUnderutilized
    consolidateAfter: 1m   # consolida subito i nodi sottoutilizzati
  limits:
    cpu: "1000"            # max 1000 vCPU totali
    memory: "4000Gi"       # max 4TB RAM totale
---
# Configurazione spot interruption handler
# Necessario per gestire le interruzioni spot (AWS invia 2 min di preavviso)
helm install aws-node-termination-handler \
  eks/aws-node-termination-handler \
  --namespace kube-system \
  --set enableSqsTerminationDraining=true \
  --set queueURL=https://sqs.eu-west-1.amazonaws.com/123456789/NodeTerminationHandler

Savings Analysis with Karpenter

# Verifica quale percentuale dei nodi sono spot
kubectl get nodes -o json | jq '
  [.items[] |
   {type: .metadata.labels["karpenter.sh/capacity-type"],
    instance: .metadata.labels["node.kubernetes.io/instance-type"]}
  ] |
  group_by(.type) |
  map({type: .[0].type, count: length})'

# Output esempio:
# [{"type": "on-demand", "count": 3}, {"type": "spot", "count": 12}]
# 80% dei nodi sono spot -> risparmio medio 65% = -$2400/mese

# Kubecost: visualizza costo spot vs on-demand storicamente
# Vai in Cost Allocation > Filter by node type

Namespace Budget: ResourceQuota as a FinOps Tool

ResourceQuotas aren't just for technical isolation — they're also a tool of financial governance. By assigning quotas based on budgets, you create a system of incentives that push teams to optimize:

# budget-quota-team.yaml
# Calcola le quote basandoti sul budget mensile del team
# Budget: 500 EUR/mese (circa 1500 CPU-hours a 0.033 EUR/CPU-hour)
# Assumendo utilizzo medio 50%: requests max = 2000 CPU-hours / 720 ore = 2.8 vCPU mean

apiVersion: v1
kind: ResourceQuota
metadata:
  name: team-alpha-budget-quota
  namespace: team-alpha
  annotations:
    finops/monthly-budget-eur: "500"
    finops/last-reviewed: "2026-01-15"
    finops/owner: "alice@company.com"
spec:
  hard:
    # Basato su budget EUR/mese
    requests.cpu: "6"       # ~500 EUR/mese a tasso medio spot
    requests.memory: "12Gi" # proporzionale
    requests.storage: "200Gi"

---
# Alert quando il team si avvicina al budget
apiVersion: monitoring.coreos.com/v1
kind: PrometheusRule
metadata:
  name: finops-budget-alerts
  namespace: monitoring
spec:
  groups:
    - name: finops
      rules:
        - alert: TeamBudgetUsageHigh
          expr: |
            kubecost_namespace_allocation_cpu_cost_hourly * 24 * 30 +
            kubecost_namespace_allocation_memory_cost_hourly * 24 * 30 > 400
          for: 1h
          labels:
            severity: warning
          annotations:
            summary: "Team {{ $labels.namespace }}: costo mensile proiettato > 400 EUR"

Savings Strategies for Different Workloads

Workload type	Strategy	Expected Savings
Web API (stateless)	Spot + HPA + rightsizing VPA	50-65%
ML batch jobs	Spot + checkpoint + scale-to-zero	60-70%
Databases (stateful)	On-demand reserved + rightsizing	20-30%
CI/CD runners	Spot + scale-to-zero (KEDA)	70-80%
Critical services 24/7	Reserved instances + rightsizing	30-40%

CI/CD Job Optimization with KEDA + Spot

# Runners CI/CD sono idle la maggior parte del tempo:
# scala a 0 quando non ci sono job, usa spot

# KEDA ScaledJob per runner GitHub Actions
apiVersion: keda.sh/v1alpha1
kind: ScaledJob
metadata:
  name: github-actions-runner
  namespace: actions-runners
spec:
  jobTargetRef:
    template:
      spec:
        containers:
          - name: runner
            image: summerwind/actions-runner:latest
            resources:
              limits:
                cpu: "2"
                memory: "4Gi"
        nodeSelector:
          karpenter.sh/capacity-type: spot  # usa spot per runners
        tolerations:
          - key: "spot"
            operator: "Exists"
            effect: "NoSchedule"
  pollingInterval: 30
  minReplicaCount: 0    # scala a 0 quando nessun job
  maxReplicaCount: 20   # massimo 20 runner paralleli
  triggers:
    - type: github-runner
      metadata:
        owner: "myorg"
        repos: "myrepo"
        targetWorkflowQueueLength: "1"

FinOps Dashboard with Grafana

# PromQL queries per dashboard FinOps Kubernetes

# Costo orario totale del cluster
sum(
  kube_pod_container_resource_requests{resource="cpu"} * 0.033 +
  kube_pod_container_resource_requests{resource="memory"} / 1073741824 * 0.004
) by (namespace)

# Efficienza CPU per namespace (utilizzo reale / requested)
sum(rate(container_cpu_usage_seconds_total[5m])) by (namespace) /
sum(kube_pod_container_resource_requests{resource="cpu"}) by (namespace) * 100

# Pod con requests molto piu alte del consumo reale (spreco > 70%)
(
  kube_pod_container_resource_requests{resource="cpu"} -
  rate(container_cpu_usage_seconds_total[24h])
) / kube_pod_container_resource_requests{resource="cpu"} > 0.7

# Importa il dashboard Grafana Kubecost: ID 11270

FinOps Best Practices for Kubernetes

Monthly FinOps Checklist

Review VPA recommendations: Analyze VPA recommendations and apply rightsizing to the Top 10 wastes
Check idle nodes: Nodes with CPU and memory usage < 10% for 24 hours must be terminated (Karpenter does this automatically)
Check spot coverage: The goal is 70-80% of non-stateful nodes on spot instances
Chargeback reports: Send monthly cost report per team via Kubecost API → email or Slack
Review ResourceQuota: Increase quotas only with business justification, not by default
Scale-to-zero for dev environments: Development environments should be shut down outside business hours (-65% cost)

Conclusions and Next Steps

FinOps for Kubernetes is not a one-time optimization: it is an ongoing process that requires visibility (Kubecost), intelligent recommendations (VPA), efficient provisioning (Karpenter with commercial) and organizational governance (ResourceQuota as budget). Implementing all these techniques together, organizations typically achieve savings of 35-55% on your Kubernetes cloud bill in the first 3-6 months.

The next step is to integrate FinOps Kubernetes into the development cycle: the developers they should see the estimated cost of their deployment before merging into production — a "shift-left" approach applied to costs.