Roadmap · Specialise
Cloud cost and FinOps (Google Cloud)
Opens with creditsFor cloud and DevOps engineers, analysts and students who want to understand, explain and reduce a cloud bill on AWS, Azure or Google Cloud, including the fast-growing AI and GPU part of it. By the end you can read a bill line by line, allocate it to teams, buy commitments safely, cut waste, forecast and catch spikes, and run a monthly cost review that engineers and finance both trust.
By the end, you will be able to
- Explain how your cloud charges, read a bill and a detailed billing export, and answer "what did this team spend and why" with a query
- Write and enforce a tagging policy, allocate shared costs fairly, and report unit costs such as cost per customer or per request
- Choose between reservations, savings plans, committed use discounts and spot capacity, and size a commitment from coverage and utilisation
- Find and fix waste: rightsizing, idle resources, storage tiers, data transfer and Kubernetes cost with OpenCost
- Forecast spend, set budgets and anomaly alerts that reach the right owner, and investigate a spike to its root cause
- Estimate and control AI spend: tokens, caching, batch, provisioned throughput and GPU capacity
- Use the FinOps Framework and the FOCUS billing format to talk to finance, and run a monthly cost review end to end
The path
9 stages, in the order to learn them. Open one to start.
1How cloud billing worksLearn what Google Cloud actually charges for: the pricing units behind every line, how projects roll up into a Cloud Billing account, list versus effective cost, actual versus amortised cost, the lines nobody estimates and why cost data is always a little late.7 topics · ~2 weeks
- 1What you are billed for: time, quantity and requestsAlmost every cloud charge is a quantity multiplied by a unit price.
- 2Cloud Billing accounts and projectsOn Google Cloud, resources live in projects, and each project is linked to one Cloud Billing account that pays for it.
- 3List, discounted and effective costThe same hour of compute can carry several prices.
- 4Actual versus amortised costIf you pay upfront for a one-year commitment, the actual (cash) view shows a large charge in the month you bought it and nothing after.
- 5The lines nobody estimatedTeams estimate compute and forget the rest.
- 6Billing data is late and keeps changingCost data is not real time.
- 7Estimating before you buildEach provider has a free pricing calculator: the AWS Pricing Calculator, the Azure pricing calculator and the Google Cloud pricing calculator.
7 in this stageOpen this stage2Reading the bill and taggingGet fluent in Cloud Billing reports and the billing export to BigQuery, then make the bill say who spent what: labels, a written tagging policy, enforcement in your infrastructure code, and a plan for spend that cannot be labelled.7 topics · ~2 weeks
- 1Cloud Billing reportsBilling reports in the Google Cloud console show cost for a billing account over a time range, grouped by project, service, SKU, location or label, with…
- 2Billing export to BigQueryCloud Billing can export cost data to BigQuery automatically.
- 3Tags and labels for costTags (labels on Google Cloud) are key-value pairs on resources, such as team=payments or env=prod.
- 4Writing a tagging policyA tagging policy is a short written agreement: the required keys, the allowed values for each, who owns each value list, where tags are applied (ideally in…
- 5Labels in billing dataOn Google Cloud, labels are the key-value pairs that appear in billing reports and the BigQuery export, and project labels apply to all costs in that project.
- 6Untaggable and shared spendSome cost can never carry your tags: support plans, taxes, some data transfer, marketplace charges, commitments bought centrally, and shared platforms such as…
- 7Your first billing queryOnce billing data is in a table, SQL answers questions the console cannot: cost by tag and service for the last 90 days, the top 20 resources by cost, or…
7 in this stageOpen this stage- Visibility and allocation: showback, chargeback and unit costsTurn a labelled Google Cloud bill into numbers each team owns: allocate by project first, split shared costs in a BigQuery view, choose showback or chargeback, and connect spend to the business with unit costs.7 topics · ~2 weeksOpens with credits
- Rate optimisation: commitments and spotPay less for the same Google Cloud usage: resource-based and spend-based committed use discounts, sustained use discounts, coverage and utilisation, Spot VMs, and why you cut waste before you commit.6 topics · ~2 weeksOpens with credits
- Usage optimisation: rightsizing, waste, storage, data transfer and KubernetesUse less on Google Cloud: rightsize with Recommender and the FinOps hub, remove idle and orphaned resources, schedule non-production, use Cloud Storage classes and Autoclass, cut data transfer cost, and see GKE cost with cost allocation and OpenCost.9 topics · ~3 weeksOpens with credits
- Forecasting, budgets and anomaly alertsSee Google Cloud spend coming: forecast with trends and drivers, set Cloud Billing budgets with Pub/Sub notifications, use cost anomaly detection, route alerts to owners, and investigate a spike to its cause.6 topics · ~2 weeksOpens with credits
- AI and GPU spendManage the fastest-growing part of most bills in 2026: per-token pricing, caching and batch, the cheapest model that passes, Vertex AI consumption options and Provisioned Throughput, GPU utilisation, calendar mode reservations and the unit cost of an AI feature.8 topics · ~2 weeksOpens with credits
- The FinOps Framework and the FOCUS billing specLearn the shared language of the field: the FinOps Foundation's principles, phases, domains and capabilities, the personas who take part, crawl-walk-run maturity, and FOCUS, the common format all three big clouds can now export.8 topics · ~2 weeksOpens with credits
- Working with finance and a FinOps careerSpeak finance, run the monthly cost review, get engineers to act, and plan a career: FinOps roles, the FinOps Certified Practitioner exam, where demand is in 2026, and a portfolio that proves you can save money.8 topics · ~2 weeksOpens with credits
7 more stages, 52 more topics, free with an account.
You have seen the open stages of the Cloud cost and FinOps roadmap. Every topic it covers is listed above, and an account opens the rest: what each topic means, a question to check yourself, the courses, the practice and Tauseef's advice for every stage. Early access is open: sign up and your account is active shortly after.
Everything here is free to learn from. A few things, like adding a roadmap, keeping a cheat sheet or asking Tauseef a question, use credits. You get 15 when you join, your first roadmap is free, and if you are learning seriously and run out, you can ask for more. Credits only make sure the platform goes to people who are actually learning.
Build a project
A finished project proves more than a certificate.
- Start with this oneintermediate2 semesters · 2 to 3 studentsMulti-Provider Cloud Bill Tracker with Anomaly Alerts for Startups
The core FinOps loop in one build: ingest billing data, allocate it by team, forecast and alert on spikes
- Python
- PostgreSQL or DuckDB
- AWS Cost and Usage exports
- Provider billing APIs and CSVs
- FastAPI
- intermediate1 semester · 2 to 3 studentsStudent Cloud Sandboxes with Budgets and Automatic Cleanup for University Labs
Budgets, tags, alerts and safe automatic cleanup for a real audience: university labs
- Terraform
- AWS, Azure or Google Cloud
- AWS Organizations or Azure Policy
- Python
- Serverless functions
- intermediate1 to 2 semesters · 2 studentsLog Volume Cost Reducer With Pattern Mining for Self-Hosted Loki and Vector Pipelines
Attacks one of the quietest large lines on a bill, logging, with measurable before and after numbers
- Python
- Drain3 log parser
- Vector or Fluent Bit
- Grafana Loki
- ClickHouse or DuckDB
- intermediate1 semester · 2 studentsAI API Spend Attribution per Feature and Customer with Budget Guards
Practises the AI spend stage on a real problem: cost per feature and per customer for AI API calls, with budget guards that stop a runaway loop
- TypeScript and Python SDK wrappers
- OpenTelemetry
- ClickHouse or PostgreSQL
- Grafana
- Redis
Related on Sefism
Picked from the other libraries by shared topics.
Best practices
See all- Kubernetes Best PracticesRun workloads on Kubernetes that schedule well, survive restarts and node drains, and stay locked down: resources, probes, PDBs, Pod Security and RBAC.
- Terraform Best PracticesWrite Terraform that teams can review and change safely: file layout, version pinning, remote state with locking, secrets, modules, refactoring and tests.
Checklists
See all- Cloud Cost Optimization ChecklistFinOps for engineers: tagging and allocation, budgets and anomaly alerts, rightsizing, scheduling, commitments, storage lifecycle, egress, idle resources and unit cost metrics.
- AWS Account Security ChecklistLock down an AWS account: root user, IAM Identity Center, roles instead of keys, CloudTrail, GuardDuty, S3 Block Public Access, IMDSv2 and guardrails.
- Incident Response ChecklistPrepare for and handle production incidents: on-call, severity levels, declaring, roles, mitigation, security containment, status updates and blameless postmortems.
Project structures
See all- Terraform project structureAn infrastructure repository for a small app on AWS: reusable modules, one folder per environment with its own remote state, pinned providers, tests and security scans, with plan on every pull request and apply from CI. For students who have finished the Terraform tutorials and want to know how a real team lays out the files. The same layout works unchanged with OpenTofu.
- Kubernetes with Helm project structureA Helm chart repository for one app: the chart that
helm creategives you, grown into a tested, versioned package with per-environment values, unit tests, install tests in a kind cluster and releases to an OCI registry. For students who can write a Deployment by hand and want to know how a team organises charts. It is about the repo layout and what each file does, not about cluster design. - Firebase (Cloud Functions and security rules) project structureFor students building an app on Firebase who need a real backend: Cloud Functions (2nd gen) in TypeScript, Firestore and Storage security rules with tests, and the Local Emulator Suite. The tree starts from
firebase initand grows into feature folders of callables, triggers and scheduled jobs, then adds CI deploys, environments, contract tests and runbooks.
Career questions
See all- Which cloud should I learn first: AWS, Azure or GCP?If you have no reason to choose otherwise, start with AWS: it has the largest market share and the most learning material. Choose Azure if the employers you want run on Microsoft, and GCP if you are heading into data and ML work at companies that use it. The concepts transfer, so the first cloud matters less than learning it properly.
- How do I practise on AWS, Azure or Google Cloud without getting a surprise bill?Use the free offer that fits you, set a budget alert before you create anything, tag every resource, build and destroy your labs with code, and run a five-minute check for the usual culprits, such as NAT gateways, load balancers, databases and public IPs, before you close your laptop.
- Are certifications like AWS or Google Cloud worth it for getting a tech job?It depends on the role. For cloud, DevOps, security and some data roles, a respected certification can help you get shortlisted, especially combined with a real project. For general software engineering roles, projects and skills matter far more. Do the certification to structure your learning, not as a replacement for building.
Cheat sheets
See allStuck on something specific?
Writing only gets you so far. If you want an answer to your situation rather than the general case, book a session and we will work through it together. Sessions are free for approved Sefism members, and a few slots open each week.
- 1:1 Mentorship
- Fresh Graduate Mentorship
- Career Guidance
Brand deals, promos and partnerships: tauseeffayyaz@gmail.com