mission-finops --what-i-do
I investigate AWS cost movement and improve the engineering behind the bill.
Independent engineering practice. Built on hands-on infrastructure experience across networks, firewalls, data centres, application migrations, DDoS mitigation, and scaling systems under pressure. Then nearly seven years at AWS advising large enterprise customers running complex multi-account estates. Now helping engineering teams investigate and improve the systems behind their cloud bill.
If the AWS bill goes up, I help find out what changed.
If the system is wasteful, I work with engineers to improve it.
If finance and engineering are speaking past each other, I help both sides work from the same evidence.
is-this-you
Is this you?
- The migration finished and the bill is double the business case.
- Spend moved and nobody agrees why.
- You have optimization recommendations your engineers refuse to implement.
- Finance and engineering produce different numbers from the same account.
Bring the messy question. That is the whole intake.
mission-finops --pricing
Under $50,000 a month on AWS?
The investigation is free. Not a trial, not a teaser. The same fixed-scope investigation, the same written explanation, the same evidence.
Above $50,000 a month, it is $5,000 USD fixed, half on start and half on delivery.
start-here
Most engagements begin with an AWS Cost Investigation. One material cost question, investigated against your billing data and your architecture. You get the evidence, the confidence limits, the likely owner, and the next actions. Free under $50,000 a month on AWS. $5,000 USD fixed above that. Roughly two weeks either way.
Read-only AWS access. Published IAM policy.
Runs locally. No mandatory raw billing-data upload.
None. No phone-home. No usage tracking.
Stays with you. No SaaS. No mandatory upload.
Open source. Inspect before you trust.
I publish my own defects. Trust ledger.
This website (not the CLI) records two events: a booking-link click, a sample-report view. No cookie. Privacy.
The analysis can remain inside your environment. There is no mandatory raw billing-data upload, and export-based engagements are explicitly scoped. Your billing data stays under your control.
mission-finops --what
Cloud cost is the result of engineering decisions.
A bill is where the number appears. The cause lives somewhere else: architecture, usage, deployment choices, ownership, commitments, data movement, or a system that no longer fits the job.
- bill-changed
I investigate the movement and connect it to the workload, service, account, or business event that caused it.
- recs-pile-up
I help engineers decide what is worth changing, what is safe to leave, and what creates a better system.
- finops-gap
Evidence, queries, and runbooks from the investigation stay with your team, in a form they will actually use.
- data-control
I work local-first and read-only. No sensitive business context leaves your environment.
When the bill misses forecast, finance has to explain a number they did not create.
The explanation has to survive scrutiny from people who will not accept "cloud went up." It needs attribution to a workload, a decision, or a change. It needs confidence limits. It needs to hold when someone asks a follow-up question in the next budget review.
That is what an investigation produces: documented evidence connecting variance to cause, with enough technical grounding that engineering can confirm it and enough clarity that finance can defend it.
mission-finops --how
I work with your builders, not around them.
I sit with the people who know the system, test the cost story against technical reality, and help turn the answer into better engineering.
Some of the useful work becomes open source. Some stays inside your company. Either way, your team owns more of the capability when I leave.
mission-finops --proof
Kulshan is the evidence engine behind the work.
Free, open-source AWS cost investigation CLI. Runs locally with read-only access and produces evidence-backed reports from the customer's own environment.v0.6.1APACHE-2.0
Network spend increase +34% m/m
14 idle EBS volumes ($312/mo)
NAT Gateway anomaly (z-score 2.7)
VPC Flow Logs not enabled
$ kulshan report --packs cost,sweep,topo --days 90 --format json -y kulshan 0.6.1 | cost, sweep, topo packs | 90 days ⬢ CUR/Data Export found ───────────────────────────────────────────── human_review_required: yes report written → ./kulshan-report.json
Not a subscription. Not a dashboard. Proof of the way I work: keep data controlled, make the method inspectable, give the builder something useful. Source is on GitHub, the IAM policy is published, and the full report shape is on the sample page (source of truth for these figures).
whoami
Mission FinOps came from engineering, not finance.
I am Yuvdeep Singh. Nearly seven years inside AWS, which is how I know what a migration plan promises, what it usually leaves out, and where the business case quietly stops matching the invoice. I do not work there anymore, and that is the part that matters. There is no partner quota behind anything I tell you.
At AWS, that work ran from the engineers operating the environment to operational reviews with CIOs, CTOs, CFOs and other executive stakeholders, covering technical health, reliability, risk, architecture and cloud economics.
Before AWS: networks, firewalls, data centres, architecture leadership across complex on-prem environments.
That path is why Mission FinOps does not treat cost as a finance-only exercise. I can trace the engineering deeply enough to produce an answer that survives the executive meeting. More on my background.
mission-finops --engage
Bring me the messy question.
"The bill moved and nobody agrees why" is enough. So is "we know there is waste but the engineering team does not trust the recommendations."
I scope engagements after a free 30-minute engineering review. See how the work is shaped.
calendar → book