Hire a Software Development Team in India — and Actually Own It
Instead of outsourcing your development, you own the team. We handle everything that isn't the work itself.
Your engineers in India sit in your sprints, your standups and your repo. You set the backlog, you approve the pull requests, you decide what ships. We recruit them, employ them, pay them, keep them, and replace them when we have to. Nobody else gets their hours.
Owning a team is a different thing from outsourcing
The word outsourcing describes handing a problem to somebody else and getting a result back. You write a spec, you agree a price, a vendor disappears for eleven weeks, and something arrives that is roughly what you asked for eleven weeks ago. Change requests cost money. The people who wrote the code are strangers. When the contract ends, the knowledge walks out with them.
This is the opposite arrangement. You are not buying a deliverable. You are building a unit of engineers who work only for you, in your systems, on your priorities, for as long as you want them. The model has an unglamorous name in the industry: an offshore development centre, or a captive unit. It is what 506 of the Forbes Global 2000 have quietly been doing in Bengaluru and Hyderabad for twenty years. The only reason a 20-person company cannot do it alone is the setup: an entity, a payroll, a compliance function, a recruiter, an office, an IT desk. That part is ours.
The difference shows up in the boring places. Who chooses the ticket that gets worked on Tuesday morning. Who runs the retro. Whether the engineer who fixed the payment bug in March is still around in September to fix its cousin.
| Owned team in India | Project outsourcing | Freelance marketplace | |
|---|---|---|---|
| Who sets priorities | You, every morning | The signed scope document | Whoever pays the most that week |
| Who runs standup | Your team lead or your EM | Nobody you meet | There is no standup |
| Code review | Inside your repo, your rules | Delivered at milestones | Ad hoc, if at all |
| Exclusivity | Full time, only you | Shared across accounts | Splitting time across clients |
| Knowledge after 12 months | Sits with people still on your team | Leaves with the vendor | Gone with the contract |
| Changing direction | Free — reprioritise the board | Change request and a new quote | Renegotiate or rehire |
| Who carries HR and payroll | We do | The vendor | Nobody |
| Cost visibility | Fixed monthly rate per seat | Quoted per project | Hourly, variable, uncapped |
The five roles in a real team, and when you actually need each one
Most vendors will sell you five developers and call it a team. Five developers is not a team, it is five developers. What turns them into a team is somebody whose job is the quality of the work rather than the volume of it, and eventually somebody whose job is the people. Here is the honest version of when each role earns its salary.
Junior developer
1–3 years $960/moWrites well-specified tickets under review. Fixes bugs, builds CRUD screens, writes tests, handles the migration nobody senior wants. A junior is a bet: cheap now, valuable in eighteen months if somebody teaches them.
When to add one: Add one once you have at least one senior with the patience to review their work daily. A junior with nobody reviewing them is the most expensive person on the team, because you pay twice. Once for the code, again for the rewrite.
Mid-level developer
3–5 years $1,400/moThe engine of the team. Takes a ticket described in two sentences and comes back with a working feature and reasonable tests. Does not need the architecture explained twice. Will ask before making a decision that is hard to undo.
When to add one: This is the role you hire most of. A team that is all senior is expensive and argumentative; a team that is all junior does not ship. Mid-level engineers do the bulk of the output on almost every product I have seen staffed well.
Senior developer
5+ years $1,800/moOwns a domain. Designs the thing before building it, spots the failure mode in your schema, and can be handed a vague problem instead of a ticket. Reviews other people's code and makes it better without making enemies.
When to add one: You need one from day one. A pod with no senior is a pod that will produce something that works in demo and falls over at 400 concurrent users. If you only budget for one experienced person, make it this one.
Team lead
7+ years, still writes code $2,400/moRuns the sprint. Breaks the roadmap into tickets, unblocks people, chases the flaky test, runs standup and retro, and still ships perhaps 40% of a developer's output themselves. The buffer between your intent and the team's day.
When to add one: The trigger is headcount, not seniority. Somewhere around four to six engineers, whoever has been informally coordinating stops having time to code. That is the moment. Below four people you do not need one and should not pay for one.
Engineering manager
10+ years, mostly out of the editor $3,200/moOwns the people and the delivery, not the tickets. Hiring, one-to-ones, performance conversations, career paths, capacity planning, and the quarterly conversation with you about what the team can realistically absorb. Manages leads, not individuals.
When to add one: Around ten to twelve engineers, or the moment you have two pods running in parallel. One team lead can hold six people. Two leads with nobody above them will quietly optimise for different things and you will find out six months later.
All rates are full-time, 160 hours a month, and include salary, employer taxes, workspace, equipment, HR and our management fee. There is no recruitment fee and no setup charge. Individual specialist seats are listed on the full rate card.
What a team actually costs, in three real shapes
Rate cards are easy to publish and hard to use, because nobody hires "a developer". You hire a mix. These are three shapes we build most often, with the seniority split and the number at the bottom. Every figure is list price. Nothing is hidden below it.
Starter pod
You have product-market fit and a founder or CTO who can hand out work directly. No lead needed yet, because you are the lead.
- 1 × Senior developer$1,800
- 2 × Mid-level developer$2,800
What it covers: Ships one product surface. Good for a v2, a mobile app alongside a web app, or taking a prototype to production.
Squad
The most common shape, and the first one that runs itself. The lead owns the sprint so you stop being a bottleneck at 8am.
- 1 × Team lead$2,400
- 2 × Senior developer$3,600
- 2 × Mid-level developer$2,800
- 1 × Junior developer$960
What it covers: Owns a whole product area end to end, including its on-call. Enough slack to absorb one person being on leave.
Department
Two pods, two leads, an engineering manager above them. This is a functioning offshore engineering organisation, not a staffing arrangement.
- 1 × Engineering manager$3,200
- 2 × Team lead$4,800
- 3 × Senior developer$5,400
- 4 × Mid-level developer$5,600
- 2 × Junior developer$1,920
What it covers: Runs a platform plus a product stream, or two product streams. Handles its own hiring pipeline with our recruiters.
The comparison that matters to your board
The six-person squad above runs $117,120 a year. Fully loaded, meaning salary plus payroll taxes, benefits, equipment, desk and recruiter fees, that is roughly what one mid-level engineer costs in San Francisco, or about two in Manchester. Whether that trade is right depends entirely on the next four sections, which is why they are on this page and not on anybody else's.
Volume pricing starts at ten seats. Below that, the numbers above are the numbers. If you want the same maths run against your own current cost per engineer, ask Rita and she will do it in the chat window rather than sending you a PDF next Thursday.
Why build a development team in India rather than anywhere else?
The strongest argument is not cost, and it is not the size of the talent pool. It is that the exact structure described on this page is already the dominant way large companies staff engineering in India, and it has been stress-tested by people with far more to lose than you.
India hosts 2,117 global capability centres employing 2.36 million people, and 506 of the Forbes Global 2000 run one (Nasscom-Zinnov India GCC Landscape Report 2026, data as at March 2026). A capability centre is an owned offshore team. Microsoft has more than 22,000 people in India; SAP Labs India runs 40% of SAP global R&D; Adobe India, at over 8,000, is Adobe largest workforce outside the United States. What we do is the same structure without you incorporating an Indian subsidiary to get it.
The pipeline argument, stated accurately
You will read on competitor pages that India produces 1.5 million engineers a year. That number is wrong by roughly double. It is India approved undergraduate engineering seat count, not its graduate count, and repeating it is a quick way to lose a technical buyer who checks.
The accurate version is still enormous. India graduated about 834,000 engineers at bachelor level in 2023-24, and 1.79 million students are enrolled in computer engineering degrees right now (All India Survey on Higher Education 2023-24, Ministry of Education). For a team of six, what matters is not the total anyway. It is how many people clear your bar, which brings us to the part nobody publishes.
The catch, part one: quality variance is real, and it is mostly a sorting problem
The statistic thrown at this industry is that only about 18% of Indian engineers are employable as software engineers. It comes from the Aspiring Minds National Employability Report 2016, and it measured students who graduated in 2015 against one specific corporate role. Other rows in the same table say 3.67% and 40.57%, so anybody quoting a single number is quoting selectively.
Take it at face value anyway, because the honest reading helps you. Roughly 18% of around 834,000 graduates is still on the order of 149,000 people a year clearing a benchmarked software bar. The problem was never supply. The same research found that employability varies enormously by college and that a large share of capable candidates study outside the top 75 institutions, while recruiters use the college name as their main shortlisting signal.
That mispricing is the entire commercial case for using somebody who screens properly. We interview against a defined rubric with the same questions in the same order, scored consistently, because structured interviews are the best-evidenced selection method available (Sackett, Zhang, Berry and Lievens, Journal of Applied Psychology, 2022). If you hire on brand names you will pay more for the same engineer. If you hire on assessment you will not.
The catch, part two: attrition, and the number that reframes it
Indian technology attrition was genuinely alarming four years ago. Cognizant trailing-twelve-month voluntary attrition peaked at 32% in mid-2022; Infosys hit 27.7%. If your impression of India was formed then, it was accurate then.
It is not accurate now. The five largest Indian IT employers currently report attrition between 12.3% and 13.8%: Infosys 12.6% for the year to March 2026, Cognizant 12.3% in Q1 2026, Wipro 13.8% in Q4 FY26, all from their own SEC filings. That tight clustering across five companies is the signal.
Now the comparison nobody in this industry publishes. The US Information sector voluntary quit rate in 2025 was 15.3% (US Bureau of Labor Statistics, JOLTS). Both figures are voluntary departures, both are current. On the present data an Indian engineering team is marginally more stable than an American one. The measures are close rather than identical, since JOLTS counts establishments and Indian attrition is reported per firm, but the direction is not in doubt.
The mitigation is structural rather than motivational. Indian notice periods run long, commonly 30 to 90 days for senior roles against statutory minimums of about 30. That cuts both ways honestly: it slows your start date, and it hands you two to three months of warning and handover that at-will US employment never gives you. Slow to start, slow to lose.
Where your code and your data actually stand
Two things engineering leaders ask about, and one they should ask about but usually do not.
India has no European adequacy decision, so EU personal data reaches India under the European Commission Standard Contractual Clauses. For UK data you need the ICO International Data Transfer Agreement, or the EU clauses plus the UK Addendum, because the EU clauses alone are not valid for UK transfers. A transfer risk assessment goes with them. India also has its own statute now: the Digital Personal Data Protection Act 2023, with rules notified in November 2025 and an eighteen-month phased compliance window.
The thing nobody mentions is section 72A of the Indian Information Technology Act 2000. It makes it a criminal offence for a service provider to disclose information obtained under a lawful contract, so the exposure runs beyond a civil claim for breach. Very few competitor pages know it exists, and it is a stronger backstop for your source code than any NDA on its own.
And the objection you may already be holding: India sits on the United States Trade Representative Priority Watch List for intellectual property, most recently in April 2026. Read the findings and they concern pharmaceutical patents, counterfeit goods and copyright piracy. They are not about source-code confidentiality or trade secrets in service contracts. Worth knowing before somebody raises it in your risk review, and worth knowing what the answer is.
How do the time zones actually work?
India runs on IST, which is UTC+5:30. That half hour is not a rounding error, it is the thing that makes the arithmetic annoying, so here is the arithmetic done for you. A standard India day is 9:00am to 6:00pm IST. We also run two shifted patterns, and the shift you pick decides whether your standup is a conversation or a written summary.
| India shift (IST) | London | New York | San Francisco | Sydney |
|---|---|---|---|---|
| 9:00am – 6:00pm | 3.5 hrs overlap | none | none | 3.5 hrs overlap |
| 12:30pm – 9:30pm | 7 hrs overlap | 2 hrs overlap | none | none |
| 2:30pm – 11:30pm | 8 hrs overlap | 4 hrs overlap | 1 hr overlap | none |
Overlap measured against a 9:00am to 5:00pm working day in each city, on winter clocks.
Read the top row again, because it is the sentence the rest of this industry avoids. On a strict nine-to-five at both ends, India overlap with the United States is exactly zero, East Coast and West Coast alike. Every hour of US overlap that exists anywhere in offshore delivery was manufactured by somebody agreeing to work outside normal hours.
So the question is never how much overlap you get. It is who gives up their evening and whether they are paid for it. On our teams that person is the engineer, the shift is agreed before they accept the offer rather than imposed afterwards, and a late shift carries a premium built into the rate you were quoted. If a vendor shows you four hours of US overlap and no shift premium, ask who is absorbing it.
One operational detail worth having: India does not observe daylight saving, so the gap to every client market moves twice a year. Worse, the US springs forward on the second Sunday in March and the UK on the last, so for two to three weeks each spring your team offset to New York has changed and its offset to London has not. A team serving both markets re-bases its roster four times a year, not two.
London and Dublin
The easy one. A 12:30pm IST start gives you seven hours together, which is a normal working relationship with a slightly early finish. Nobody has to compromise. If your team is in the UK, stop reading this section.
New York, Toronto, Austin
A 2:30pm IST start buys four hours, 9:00am to 1:00pm your time. That is enough for a live standup at 9:15, a design conversation before lunch, and pairing when something is on fire. Your afternoon is their evening, so anything you send after 2pm gets picked up tomorrow.
San Francisco, Seattle, Vancouver
This is the hard one and we will not pretend otherwise. Four hours of overlap with US Pacific requires an India shift running roughly 10:30pm to 7:30am IST. People do it, and people burn out doing it. See the section on what goes wrong below.
What stays yours, and what becomes our problem
The cleanest way to explain this model is to draw the line and then never cross it. Everything on the left is a decision only you should be making. Everything on the right is administration that would cost you a country manager, a payroll provider and a lease to do yourself.
You own
The work and everything that shapes it
- ✓The backlog, the roadmap and what gets built this sprint
- ✓Sprint length, ceremonies and your definition of done
- ✓Code review standards, branching model, merge rights
- ✓Architecture and technology choices
- ✓Who on the team does what, and who is on call
- ✓Hiring decisions — you interview and you say yes or no
- ✓Performance feedback, delivered through your lead or EM
- ✓All source code, IP and data, assigned to you in writing
We own
Everything that is not the work
- →Sourcing and screening candidates against your brief
- →Technical assessment before anyone reaches your interview
- →Background and reference checks
- →Employment contracts, offer letters, notice periods
- →Payroll, PF, gratuity, income tax and statutory filings
- →Health insurance, leave policy, appraisal cycles
- →Office space, desks, machines, dual monitors, backup power
- →Redundant internet, VPN, endpoint security, device encryption
- →Retention — the awkward conversation before somebody resigns
- →Backfill recruitment when somebody leaves anyway
What the first ninety days look like
Nobody is productive in week one, and a vendor who tells you otherwise is describing a sales deck rather than a team. This is the timeline we actually plan against, and the thing we ask you to judge us on at each point.
You describe the roles, the stack, the shift and the seniority mix. Within 48 business hours you get two to three CVs per seat, each already through a technical screen. You are reading real profiles, not a capability statement.
You interview. You reject people. Most clients reject at least one, and that is healthy. Offers go out the same day you decide. Indian notice periods run 30 to 90 days for people currently employed, which is the single biggest scheduling variable. We shortlist candidates on shorter notice when your start date is tight, and we tell you upfront which is which.
Accounts provisioned in your systems by you, under your SSO. Machines already imaged. The engineer reads the codebase, runs it locally, and ships one small, real change by Friday: a copy fix, a test, a dependency bump. The goal is proving the pipeline works, not proving they are fast.
Working through well-specified tickets with review on everything. Expect roughly half the output you will get later. This is when you find out whether your onboarding documentation exists, because their questions will expose every gap in it. Weekly checkpoint with us on how it is going, separately from your own reviews.
The engineer starts pushing back on tickets. They spot when a requirement contradicts something built in March. Estimates start being useful. If you hired a lead, the lead is now running the sprint and you have stopped attending every standup.
Full velocity, and the point at which we ask you to score the team honestly. If somebody is not there by the end of month three, that is a conversation we start rather than one you have to raise. Most teams add their second pod somewhere between month four and month eight.
How reporting works when nobody is in the same building
"The team reports to you" is a claim every vendor makes, so here is what it means operationally. Your engineers have accounts in your systems, provisioned by your IT under your single sign-on, and revocable by you in thirty seconds without asking us. They appear in your org chart. They have your email address, not ours.
We do not sit between you and them. There is no account manager forwarding your questions. The only standing meeting we hold is a monthly thirty minutes with you about retention, leave, and whether anybody is drifting. That is the employment conversation, not the delivery one.
Tooling
Jira, Linear, Shortcut, Azure DevOps or a Trello board. Whatever you already run. GitHub or GitLab under your organisation, with your branch protection rules. Slack or Teams, in your workspace, as full members rather than guests where your plan allows it. We do not ask you to log into a portal of ours to see what your own team did yesterday.
Ceremonies
Your sprint cadence, your standup time, your retro format. If you run two-week sprints with Thursday demos, so do they. The India shift is chosen to make your standup live rather than asynchronous, which is usually the deciding factor between a 9:00am and a 2:30pm IST start.
Escalation
Delivery problems go to your lead, then to you, the same as with any employee. Employment problems come to us: someone is unwell, someone wants to relocate, someone has a competing offer. You hear about all three early. The distinction matters: you should never be the person negotiating an Indian salary revision.
IP and confidentiality
Every engineer signs an NDA and a written IP assignment before their first login, naming your entity as the owner of everything they produce. Behind the contract sits section 72A of the Indian IT Act 2000, which makes disclosure of information obtained under a lawful contract a criminal matter rather than only a civil one. We will sign your paper instead of ours if your legal team prefers it, and most do.
The four boundary questions nobody answers
Every vendor on this market says the team works under your management. Almost none of them will tell you who does the things a manager actually does. Here is where the line falls on ours.
Your lead or your EM, weekly, about the work. We hold a separate monthly conversation about career and wellbeing, and we tell you anything from it that affects your delivery.
You do the assessment, because you are the only person who has seen the work. We run the formal appraisal cycle and deliver the salary decision, because we are the employer and Indian pay-review conventions are not something you should have to learn.
You propose, we price and execute. If you want to move a mid-level engineer to senior, that is your call on merit and ours on what the increment has to be to make it real.
You can end anybody's assignment at any time, for any reason, and we absorb it. We cannot move anybody off your team without your agreement. That asymmetry is the whole point of the arrangement.
What this costs you that is not on the invoice
Every page selling this model prices it as salary times headcount. That understates it, and pretending otherwise is how engagements get abandoned in month five with everyone feeling cheated. Three real costs sit on your side of the ledger.
The first is your overlap hours. Four hours a day with New York is four hours during which somebody senior on your side has to be reachable. That is a genuine constraint on your own calendar, and if your CTO is also your head of sales it will not hold.
The second is review load. For the first two months your onshore engineers will spend real time reviewing code and answering questions. Budget roughly a fifth of one senior person for a six-person squad, falling away sharply after month three. Teams that refuse to pay this end up with a codebase they do not recognise.
The third is documentation you have probably never written down. New engineers expose every piece of tribal knowledge your existing team carries in its head. That work was owed anyway; the offshore team just makes the bill arrive.
What happens when somebody underperforms, or leaves
Two different problems that get lumped together in most contracts. They need different handling and you should know both mechanics before you sign anything.
Somebody is not good enough
Tell us inside the first 30 days and we replace them at no cost. No debate, no justification memo. You do not pay for the replacement search and you do not pay for the overlap.
After 30 days, we still replace them, and we still do not charge a recruitment fee. But we will ask to look at three things first, in this order: whether the tickets are specified well enough for anyone to succeed, whether they have a senior reviewing their work, and whether the problem is the person. In our experience roughly one placement in twelve gets replaced, and most of those happen inside the first fortnight.
The thing we will push back on: replacing a junior for behaving like a junior. That is what the price reflects.
Somebody resigns
This will happen. Indian technology attrition is real and no vendor makes it disappear. What we control is the warning and the handover.
Standard notice on our contracts is 60 days, which is long by Western standards and deliberate. You get told the day the resignation lands, not the week before they go. Backfill sourcing starts that same day and you interview replacements while the outgoing engineer is still at their desk. Where the timing works we run a two to four week paid overlap so the handover is a person, not a document.
Structural protection matters more than any of that: no single engineer is the only person who understands a service. Your lead enforces that, and it is one of the reasons a six-person squad survives a resignation and a three-person pod feels it.
When you should not do this
We turn down roughly one enquiry in five for the reasons below. An owned offshore team is a good structure with narrow prerequisites, and forcing it where it does not fit produces an expensive failure that everybody blames on India.
You need fewer than three engineers
One or two people is not a team, it is a hire. You get the same engineers with less structure and less commitment through staff augmentation, and you should not pay for a lead you do not need.
You have a fixed scope and a fixed date
If the requirement genuinely will not change and you want somebody else to carry the delivery risk, buy a project. A monthly team is the wrong commercial shape for a one-off build with a deadline attached.
Nobody on your side can specify work
The model assumes somebody in your company can turn intent into tickets and answer questions within a day. If your product function is one overloaded founder who is in sales meetings all week, the team will idle and you will conclude they are slow.
Your whole team is on US Pacific time and you need daily pairing
Four hours of overlap with San Francisco means an India night shift. It is possible, we staff it, and the honest position is that retention on permanent night shifts is worse and the talent pool willing to do it is smaller. If your engineering culture depends on constant synchronous contact, look at Latin America instead.
You are hiring for a six-week experiment
Ramp-up costs you the first month either way. Anything under a six-month horizon does not repay the onboarding, and we would rather say that now than invoice you for it.
You want the cheapest possible number
There is always someone quoting 30% below us. They are usually shifting people between accounts, or hiring a grade below what they invoiced, or both. If price is the only variable you are optimising, you will get what you optimised for.
This page, or one of the other two?
We sell three arrangements that people confuse constantly. They differ in structure, not in the quality of the people, and picking the wrong one is the most common mistake buyers make.
A software development team
You are on this page
Build a whole structured unit with its own leadership, juniors through to an engineering manager, that exists only to work on your product, for years. You are creating an engineering organisation offshore.
Staff augmentation
Best for 1–2 people
You already have an engineering team and a manager running it. You need one Node developer and a QA engineer plugged into that existing structure next month. No new leadership, no new unit.
Plug individuals into your existing team →The dedicated team contract
The commercial model
The terms underneath all of this: exclusivity, minimum commitment, notice periods, how scaling up and down is billed, and what the monthly rate does and does not include across every role we staff.
Read the contract and commercial terms →Hiring for one specific stack rather than a whole team? The individual role pages go deeper on skills and vetting: React developers, Node.js developers, Python developers, DevOps engineers and QA engineers. Or start at the full list of engineering roles we staff.
Questions buyers ask before they commit
A three-person pod of one senior and two mid-level engineers is $4,600 a month. A six-person squad with a team lead is $9,760. A twelve-person department with an engineering manager, two leads and nine engineers is $20,920. Individual seats run $960 for a junior, $1,400 mid-level, $1,800 senior, $2,400 for a team lead and $3,200 for an engineering manager, each full-time at 160 hours a month. Those rates include salary, employer taxes, workspace, equipment, HR and management. Volume pricing begins at ten seats.
Yes, and it is written into the contract rather than promised on a call. Each engineer is assigned to one client, full-time, and is never rotated onto another account or billed to two places at once. They use your email address, appear in your org chart and attend your ceremonies. If you ever suspect otherwise, you have visibility of their commit history and their calendar, because both live in your systems and not ours.
You do. You set priorities, run the sprint, review the code and decide what ships. If you buy a team lead or an engineering manager, that person runs the day to day on your behalf and reports to you, not to us. Our involvement is employment: payroll, leave, appraisals, retention and backfill. The one standing meeting we hold with you is thirty minutes a month about the people, not the product.
Outsourcing transfers the problem. You define a scope, agree a price, and receive a result you had limited visibility of while it was being built. This transfers the employment overhead and nothing else. The engineers are yours in every practical sense. Your backlog, your repo, your standards, your call on what gets built — and we carry recruitment, payroll, compliance, infrastructure and retention. When the engagement ends, the codebase and the context stay with people you chose.
Candidates currently employed elsewhere typically serve 30 to 90 days, which is the biggest variable in your start date and the one most vendors are vague about. We label every shortlisted CV with its actual notice period so you can weigh a stronger candidate against an earlier start. Our own contracts carry a 60-day notice, which is why you get two months of warning before anybody leaves your team.
Yes, within limits worth understanding. A 2:30pm to 11:30pm IST shift gives four hours of live overlap with New York and eight with London. A standard 9:00am to 6:00pm IST day gives three and a half hours with Sydney. US Pacific is the genuinely hard case: four hours of overlap needs an India night shift, which we will staff but which narrows the candidate pool and shortens average tenure. We would rather tell you that before you sign than after.
You do, from the first commit. Every engineer signs an NDA and a written IP assignment naming your entity before they receive any access, and the assignment covers everything produced during the engagement. The code lives in your GitHub or GitLab organisation under your branch protection rules, so ownership is a matter of where the repository sits rather than a clause anybody has to enforce. Most clients ask us to sign their own agreements instead of ours, and that is fine.
Yes, and it is worth asking early because most vendors will not put it in writing. Once your team passes roughly twenty engineers, running your own subsidiary usually becomes cheaper than paying anybody a management margin, and at that point we would rather transfer the team than lose the relationship badly. There are no non-solicitation clauses preventing you from employing your own people directly. We agree a transfer fee and a handover window at the point you decide, not as a penalty buried in the original contract.
Three months to start, which reflects the ramp-up rather than a lock-in tactic. The first month is onboarding and you would be paying for it either way. After that it is rolling monthly with 30 days notice to reduce headcount. Scaling up is faster: additional engineers in the same stack are usually shortlisted within 48 hours, with the start date set by that candidate's notice period. There is no penalty either direction and no exit fee.
Tell Rita your roadmap. Get a team shape and a monthly number back.
Answer four questions in the chat. What you are building, your time zone, your current engineering headcount, and when you need people. Rita comes back with a seniority mix, the monthly cost, and CVs within two working days. No form, no discovery call you have to survive first.