Now Live AI Infrastructure Audit — Free 30-min review for SaaS & AI teams
Book Discovery Call
Home / FAQ
31 questions · honest answers

The questions buyers actually ask us.

Most FAQ pages answer marketing-team questions. This one answers the questions our buyers actually email us — before kickoff, mid-engagement, and after handover. Procurement, legal, payment terms, scope changes, IP, security. No hedging.

31
Questions answered Across 6 categories
0
Marketing fluff Every answer written by a senior engineer
47
Engagements behind it These are the questions that came up across them
i
Section i · 5 questions

Before you reach out.

Things to figure out for yourself first — so you know whether a call with us is worth thirty minutes of your time. If after reading these you think we’re not a fit, that’s a useful outcome.

Probably, if you’re a Series A–C SaaS or AI company with at least a few engineers, you’re running production infrastructure on AWS, GCP, or Azure, and the problem you’re looking to solve sits in one of our four practice areas: cloud architecture, cost optimization, reliability engineering, or AI infrastructure.

Probably not, if any of the following:

  • You’re pre-product or pre-traction (you don’t need senior infra yet)
  • You want a body shop to staff a project (we don’t do that)
  • You need on-premise / bare-metal expertise (cloud only)
  • You’re looking to outsource your entire engineering function (we’re a sprint, not a department)

If you’re not sure, ask on the first call. We’ll say no plainly and send referrals if there’s a better fit elsewhere.

Honest fit
No body shop

Our smallest engagement is the $8k Cost Optimization audit, which produces a written report with a prioritized list of savings opportunities. That’s realistic for a $10k–15k budget. Below that, we recommend doing the work yourself using our public Free Infrastructure Review checklist — same diagnostic, no fee.

If your budget is constrained but the work is substantial, tell us on the call. Sometimes we can scope the engagement smaller, defer pieces to a later phase, or recommend a more specialized (cheaper) shop.

Pricing
No-strings option

No. We don’t do free POCs to prove ourselves. The reason: free POCs distort scoping — we’d either underdeliver to fit the freebie window, or overdeliver and pretend it was free. Neither serves you.

What we do instead: the $8k Cost Optimization audit serves the same trust-building function. It’s small, finite, and at the end you have a concrete written deliverable. If you don’t want to work with us after that, you’ve still got a usable artifact.

Boundaries
$8k audit

Because the alternative — vague pricing that depends on who’s asking — is the consultancy industry’s most reliable indicator of bad-faith. Different prices for different perceived budgets is how agencies extract maximum WTP. We’d rather lose work to a clearly-priced competitor than win it through opacity.

Practical consequence: you don’t need a call to know if we’re in budget. You can read the service pages, see the numbers, and decide whether it’s worth booking time. We get fewer wasted intro calls. Buyers don’t waste their time. Both sides win.

Transparent pricing
Anti-script

Yes, and we don’t charge a placement fee. If during an engagement you decide you want to hire the engineer permanently, talk to us. We’d rather one of ours go in-house on a problem they’d been working on than make them stop and hand off cold.

What we ask: complete the engagement first. Don’t poach mid-sprint — it breaks the team and the timeline. After handover, all conversations are open. We’ve sent three engineers to client teams this way since 2022.

Hire-to-permanent
No placement fee
ii
Section ii · 5 questions

The first call & proposal.

What actually happens between “I emailed Cloudico” and “we signed the proposal.” Pure logistics, no theater.

A named senior engineer — the same engineer who would lead the engagement if you booked it. Not an SDR. Not a sales rep. Not “the principal architect who designed the scope.” The engineer’s name, role, and LinkedIn are in the calendar invite before the call.

Typically Hassan for Cloud or AI Infrastructure work, Maryam for Reliability Engineering, David for ML platform work. If multiple practice areas are in play, one of them runs lead and we’ll loop in the other for technical questions.

First call
Named engineer

Yes. Reply to the calendar invite or email us beforehand and we’ll send our standard mutual NDA. Or send yours — we’ll review and sign reasonable terms within 24 hours. Common edits we don’t fight: 3-year confidentiality, US or English law jurisdiction, exclusion of independently-developed work.

What we won’t sign: non-competes that prevent us from working in your industry, non-solicits longer than 24 months, or confidentiality terms covering “everything we hear” without bounds. We always treat the call as confidential by default regardless of paper.

NDA
Legal

If we’re a fit and you want a proposal: within 48 hours of the call, or not at all. A real written scope — not a deck — with fixed price, fixed timeline, explicit out-of-scope list, deliverables, payment schedule, and the named engineer.

If 48 hours pass without one, assume something has changed (rare, but it happens) and ping us. Sometimes the scope reveals itself to be larger or smaller than the call suggested, and we’ll write to you with a re-scope question rather than guessing.

48-hour proposal
In writing

No. If you say no, we don’t chase. No “checking in” emails. No “is this still on your roadmap?” six weeks later. No CRM nurture sequence.

The only follow-up email we send is our monthly infrastructure brief, and that requires you to explicitly subscribe. If you didn’t subscribe, you won’t hear from us again unless you reach out first. About a third of our engagements come from someone who passed initially and circled back 6–18 months later. The quiet apparently helps.

No follow-up
No CRM

No. We don’t commit to timelines we haven’t scoped. If you need a hard date by next Tuesday, the call is where we figure out whether that date is achievable. Sometimes the answer is “yes, easily.” Sometimes it’s “not by us, but here’s who could.” Sometimes it’s “achievable, but at higher cost — here’s the trade-off.”

Capacity availability: we cap at 8 concurrent engagements. Current waitlist is typically 2–4 weeks for a new sprint to start. Cost Optimization audits can usually start within a week.

Capacity
2-4 wk waitlist
iii
Section iii · 6 questions

Pricing & contracts.

The boring procurement-friendly answers: payment terms, scope changes, cancellation policies, what fixed-price actually means in practice.

It means: the price in the proposal is the price you pay. If we estimated wrong and the work takes longer, that’s our cost to absorb. We don’t come back with “the scope grew” invoices unless you’ve formally added scope (see next question).

The reason this works: we cap engagements at 8 firm-wide, which forces us to be careful with estimation. We’ve eaten about $80k in overruns across 47 engagements since 2022. That’s the cost of taking estimation seriously, and we’d rather absorb it than become an agency that nickel-and-dimes change orders.

Fixed-price
We eat overruns

Two paths, depending on size:

  • Small additions (under ~10% of original scope, a few days’ work): we absorb them. It’s not worth the paperwork for either side.
  • Larger additions: we write a scope-change addendum with explicit additional cost and timeline. Same fixed-price commitment rules apply to the new work. You can accept it or defer to a follow-up engagement.

What we won’t do: silently take on additional work and then surprise you with an invoice. Every scope change either lands in a written addendum or doesn’t happen.

Scope changes
In writing

For sprint engagements: 50% on signature, 50% on handover. Net-30 on the back half. Both invoices are issued in USD, payable by wire or by ACH for US clients.

For monthly retainers: invoiced on the 1st, payable Net-30. Cancel anytime with no notice required — if you cancel mid-month, you owe a pro-rated portion of that month and nothing further.

For the $8k Cost Optimization audit: 100% on signature. The audit is short (2–3 weeks) and split-payment isn’t worth the overhead.

We work with procurement teams. Standard MSA / SOW pattern is fine. We won’t do discounts in exchange for case study rights or other indirect compensation arrangements — just clean cash terms.

Payment terms
Net-30

Yes. One Slack message or email and the next invoice doesn’t fire. No notice period. No exit interview. No “let’s discuss what we could do differently.” If we’re not delivering value, leaving us should be frictionless.

Why we structure it this way: auto-renewing quarterly contracts are how bad consultancies hold onto clients past their welcome. If our work doesn’t justify the next month’s invoice on its own merits, we’d rather you leave than feel locked in.

Cancel anytime
No exit friction

Because hourly billing aligns incentives backwards. The longer the work takes, the more we make. Fixed-price flips that: the faster and cleaner we work, the better the engagement is for both sides.

For genuine ad-hoc needs we offer the embedded SRE retainer ($12k/mo) which gives you a senior engineer for a meaningful chunk of their week. Below that threshold, we’d rather not engage — one-hour-here, two-hours-there work is genuinely hard to do well and we’d be doing it badly.

No hourly
Aligned incentives

Yes. We’ll connect you with 1–2 past clients in a relevant practice area, with their permission. We don’t flood you with reference contacts because (a) it burns out our clients on reference calls and (b) it’s not actually persuasive — everyone’s heard the carefully-curated reference call before.

What we recommend instead: ask the reference one direct question — “what would you tell me if Cloudico wasn’t cc’d on this email?” The answers to that are usually more useful than 30 minutes of polite check-the-box conversation.

References
1-2 contacts
iv
Section iv · 5 questions

Working together.

Time zones, tools, communication cadence, who’s actually in the room, security expectations. The operational reality of how we ship.

Lahore, Pakistan (UTC+5) is home. Most of our team works from there. David, our ML platform engineer, is remote from Lisbon (UTC+1).

Practical overlap:

  • US Pacific: 7–10am PT every workday
  • US Eastern: 10am–1pm ET every workday
  • EU / UK: most of the working day
  • SEA / AU: most of the working day

Async-first by default. We can do daily live calls if your team needs them, but our preference is clear written communication that doesn’t require both parties online at the same time. Most engagements have one 30-min live sync per week.

Time zones
Async-first

We adapt to yours. Whichever combination of Slack / Linear / Notion / Jira / GitHub / GitLab / PagerDuty / Opsgenie / Datadog / Grafana your team already lives in, we work in that.

We don’t ask you to learn a Cloudico-branded portal or use our project management tool. Reasons: (a) you’d hate it, (b) the work artifacts live in your tools anyway, (c) when we leave, there’s nothing to migrate off of.

One light requirement: a shared Slack channel or equivalent for day-to-day questions. Async beats email for engineering work.

Your tools
No portal

We’re not a data processor. We don’t hold your customer data anywhere. We work in your AWS / GCP / Azure account, against your repos, with credentials you provision and revoke. That structure puts us out of scope for most SOC 2 / GDPR / HIPAA Business Associate Agreements — we’re configuring your systems, not processing data through ours.

What we will do:

  • Sign BAAs / DPAs where your legal team wants belt-and-suspenders coverage
  • Use scoped IAM roles, not root credentials — we ask for least-privilege access
  • Audit-log everything we do (your CloudTrail covers it)
  • Comply with your access-revocation policies on engagement end

What we won’t do: copy production data to our laptops. If we need realistic test data, we either work in your sandbox or use synthetic data.

Compliance
Least privilege

The lead engineer on your engagement, primarily. If they’re heads-down on code or in another meeting, you can ping the shared channel and either the second engineer on the engagement (we usually have a pair) or the founder picks it up. Median response in our shared channels is under 30 minutes during business hours.

You don’t get an account manager. We don’t have account managers. If you escalate something, it goes directly to the founder.

Direct contact
No AMs

Sometimes — depends on whether we have capacity. We cap at 8 concurrent engagements firm-wide. If you want to expand and we have a slot, we’ll add the scope via the addendum process. If we don’t, we’ll tell you straight and either queue you for the next opening or recommend a complementary team.

What we won’t do: add a junior engineer just to expand revenue. The named engineer doing your work doesn’t change mid-engagement. If we can’t expand with senior capacity, we don’t expand.

Capacity cap
No juniors
v
Section v · 5 questions

Your code & IP.

Who owns what we build, where it lives, what we’ll keep confidential, and what we’ll never reuse.

You do. Work-for-hire by default. All custom code, Terraform modules, runbooks, dashboards, and documentation we produce for your engagement become yours on creation. We don’t retain a license to your specific implementation.

Carve-outs: generic patterns and open-source components are not customer-specific work product. We bring in standard Terraform modules, our own internal scripts, and open-source libraries we’ve used before. Those don’t transfer ownership — they remain available to us for use on other engagements — but they’re also things you could have used freely yourself.

IP ownership
Work for hire

In your repository, from day one. Not on our laptops as “working copies” that we hand over at the end. Every commit goes to your GitHub / GitLab / Bitbucket org, signed with the engineer’s name and email.

Practical implication: your team can see every line of work as it happens. If you fired us tomorrow you’d lose nothing — the code is already yours. No zip-file delivery at handover. No “we’ll send the final version Friday.”

Code custody
Your repo

Yes — you own it, you can release it under any license you want. Permissive (MIT, Apache 2.0), copyleft (GPL), or keep it proprietary. We have no say.

What we ask: if you do open-source something, leave the original commit attribution intact. The engineer’s name on the commits stays. We don’t demand credit in READMEs, but we won’t rewrite git history to scrub our involvement either.

Open source OK
Keep attribution

Not without your written approval. Logo / case study rights are opt-in, not default. If you want to be publicly mentioned, we’ll write the case study together and you sign off on every word before publication. If you’d rather we never mention you, we won’t.

Of our 47 engagements, 28 are entirely confidential and we don’t name them anywhere. That’s why the website is light on logo walls. We’d rather have your trust than your logo.

Confidentiality
Opt-in only

We’ll reuse generic engineering patterns we’ve worked on for years — vLLM tuning approaches, Terraform module structures, SLO design patterns, autoscaling strategies. Those aren’t company-specific knowledge.

We won’t reuse your specific business logic, architecture decisions, performance numbers, customer data shapes, or strategic priorities. That stuff is confidential by default. If you sign an NDA, that’s the explicit guarantee on paper; our practice is the same with or without.

If you have direct competitors who are also clients, we’ll tell you. So far we’ve never had two direct competitors as concurrent clients in the same practice area — we’d decline the second one.

Competitor work
We’ll decline
vi
Section vi · 5 questions

After the engagement.

Handover, ongoing support, what to do if something breaks six months later. The post-shipping reality.

In the final week of every engagement you get:

  • Recorded walkthrough sessions of every dashboard, runbook, and deployment path we built
  • Written runbooks for the top 5–10 operational scenarios (failover, scaling, rollback, common incidents)
  • An “in case of fire” doc — one page, what to do if your monitoring lights up red
  • The “what we’d do next” memo — unsolicited written advice on the next 3–6 months. Free. Sometimes implemented by your team without us.

Goal: your team owns it after handover. We’re not building a managed services dependency. We deliver, we hand over, we step back.

Handover
Real ownership

30 days of post-handover Slack support, included. Your team will have questions while they settle in — how do I run this script in staging, what was the rationale for this decision, can we tweak this threshold. We answer those in the shared channel.

After 30 days we step out cleanly. If something serious comes up later, email us directly — we don’t bill for the first quick question on an old engagement.

30-day window
First question free

Yes. Email incident@cloudico.co with a severity tag and a senior engineer responds within 30 minutes (business hours) or within 2 hours (off-hours). This is paid — we bill incident response hourly at $400/hr with a 2-hour minimum — but there’s no retainer required.

If it’s something we built that broke through our fault, we won’t bill you for the diagnosis. If it’s a new issue or something your team changed, normal billing applies.

Incident response
$400/hr

Yes, often. About 66% of our clients come back for a second engagement within 18 months — usually on a different practice area than the first.

What returning clients get: skip the discovery call, we’ve already done it once. Proposal lands faster. Scoping is tighter because we know your stack. The same lead engineer continues if they’re available; otherwise we brief the next lead with everything from round one before the call.

Returning clients
66% come back

Yes. The $8k Cost Optimization audit doubles as a yearly health check. Some past clients run it annually — it surfaces drift, new wasteful patterns, opportunities created by new AWS / GCP features, and any architecture decisions that aged badly.

If you’d rather a custom annual review, we can scope it — usually ~$12k for a 2-week deep audit across all four practice areas with a written report. Not pitched as a recurring engagement. Most teams run it every 12–18 months, not quarterly.

Annual checkup
Not pitched as recurring

Still didn’t answer your question?

The questions on this page came from real client emails. If yours isn’t here, it’s genuinely new — and we’d like to answer it. Pick a path: