Hire Python talent in India. We handle the employment.

India produces more Python developers than any country except the United States, and by most graduate counts it has already passed it. That single fact is why you are reading this page. But the headline number hides the split that actually matters when you hire: the market divides into a services-company majority who have used Python as a scripting layer inside someone else's stack, and a product-company minority who have owned Python systems in production. The two groups interview very differently and price very differently.

The product minority is concentrated. Bengaluru holds the deepest bench, followed by Pune, Hyderabad and the NCR belt. Remote-first hiring since 2020 has spread strong people into tier-two cities, Jaipur, Indore, Kochi, Coimbatore, where salaries run 15 to 25 percent below the metro premium for the same output.

Demand has shifted too. Five years ago Python roles in India meant Django CRUD or QA scripting. In 2026 the calls we run are dominated by data engineering, ML pipelines, FastAPI microservices and LLM integration work. That shift pulled senior Python pay up faster than almost any other stack, because the same person is now bid on by product startups, GCCs and AI teams at once.

It helps to know which segment of the ecosystem your role sits in, because each one draws from a different bench. Web and API work runs on Django and FastAPI, and this is the widest bench in the country. Data engineering runs on pandas, PySpark and Airflow, concentrated in companies that grew analytics teams early. ML and LLM work is the newest and thinnest layer, pulled from the same people the data teams want, which is exactly why it prices highest. A brief that says Python developer without naming the segment gets you a shortlist averaged across all three, which serves none of them.

You are not hiring from a bottomless pool. You are hiring from a thin, well-paid layer at the top of a very large pyramid. Everything in this guide, the bands, the rubric, the process, exists to help you find that layer without paying agency markups to reach it.

Indian salaries are quoted as CTC, cost to company, in lakhs per annum. One lakh is 100,000 rupees, and 1 LPA is roughly 1,200 dollars a year at 2026 exchange rates. Here are the bands we see clearing in real offers, not the survey averages that lag the market by a year.

💰 What moves a candidate inside the band

Four premiums stack on top of the base band. Data engineering or ML pipeline experience adds 15 to 30 percent, because AI teams bid on the same people. Bengaluru and Hyderabad addresses add 15 to 25 percent over tier-two cities. A recognised product-company name on the CV adds a level of negotiating confidence more than a fixed number. And an active offer in hand adds whatever it says on it, because counter-offering is standard practice in this market.

Read the band from the top when you plan budget. A mid-level engineer you actually want, product experience, decent system design, clean communication, clears at 20 to 28 LPA in 2026, not at the 12 LPA floor. Teams that budget the floor spend three months interviewing the bottom of the pyramid and then conclude that India does not work. India works. The floor does not.

Salaries move roughly 10 percent a year at the moment, with switch increments of 30 to 50 percent still normal when a developer changes jobs. Benchmark at offer time, not from a survey published last spring. We publish live band data through the salary calculator, which is refreshed from offers that actually closed.

The offer letter says CTC. Your invoice says something larger. The gap is statutory employer cost, and pretending it does not exist is the most common budgeting mistake foreign companies make on their first India hire.

💸 The two costs that hide outside the table

FX spread is the first. Wire salaries through a regular bank and you lose 3 to 5 percent on the exchange rate without it ever appearing as a line item. On a five-person payroll that is real money every month. Settle at the RBI reference rate and the leak closes. Ask any vendor which rate they settle at before you sign anything.

The second is the misclassification contingency. Pay the same person as a contractor instead, and the invoice looks 15 percent cheaper because nobody is funding PF or gratuity. If that arrangement is ever reclassified, and full-time contractors working your hours on your repos are exactly what gets reclassified, the arrears run backwards to day one with 12 percent annual interest and damages on top. The practical exposure runs 25,000 to 40,000 dollars per head. The contractor discount is a loan, not a saving.

Compare the structures properly in the EOR versus entity calculator before you fix a budget. For most teams under 10 to 15 India hires, employment through an EOR is cheaper than either the contractor gamble or an entity of your own.

Every sourcing channel in India works. They just work at very different signal-to-noise ratios, and the noise is expensive because you pay for it in interview hours.

🚀 The channels, ranked by signal

Referrals from engineers you already trust sit at the top, as they do everywhere. Below that, targeted outbound on LinkedIn and cutting through Naukri's volume with sharp filters. GitHub activity is a genuine signal for Python specifically, library contributions, typed code, real test suites. At the bottom sit open job postings, which in India generate hundreds of applications per role within days, most from the services layer of the pyramid, each one costing screening time.

The strongest Python people in India are passive. They are employed, counter-offered annually, and reachable only through direct outreach with a specific, credible pitch. A vague JD forwarded by an agency does not move them. A message that names the problem, the stack and the band does.

The community layer is worth working too, because Python in India has one. PyCon India, the city meetups in Bengaluru and Pune, and the maintainer circles around widely used libraries are where the product minority is visible in person. A speaker list from last year's conference is a better sourcing document than most agency benches, and an engineer who gave a talk on their testing setup has already passed half your screen in public. This channel is slow, measured in relationships rather than requisitions, which is exactly why almost nobody hiring from abroad uses it and why it stays high signal. A practical shortcut: search the talk archives for the libraries your stack depends on, then write to the people who presented on them. The reply rate on that kind of message embarrasses every template a sourcing tool ever produced, because the recipient can tell in one line that a human read their work.

⚠️ The agency trap

Traditional staffing agencies solve the volume problem and introduce a margin problem: 8.33 percent of annual CTC as a one-time fee at the cheap end, or a permanent 15 to 40 percent markup on a day rate at the expensive end. Worse, most agencies screen on keyword match, not on code. You still do the real filtering, you just pay for the privilege.

Our sourcing works differently and prices differently: a screened shortlist against your brief inside nine days, 12 percent of annual CTC for junior and mid roles, 15 percent for senior, invoiced only when the hire completes 90 days. Nothing at offer. If the person does not survive the first quarter, you do not pay for them. The recruiting model is documented here if you want the mechanics.

Python's low barrier to entry is its hiring curse. Everyone lists it. The rubric below is what our technical screens run against before a candidate ever reaches your calendar. Steal it even if you never work with us.

✅ The three signals that actually predict

First, the code review exercise beats live algorithm puzzles. Reviewing a deliberately flawed pull request shows judgment, taste and communication in 30 minutes, and it cannot be rehearsed from a LeetCode playlist. Second, ask the candidate to walk through a system they built and then push on one decision. People who owned the work defend it with specifics. People who sat near the work go vague within two questions. Third, written communication. Your India team will work async with your home team. A candidate who writes clear English in the take-home summary will write clear pull requests, clear Slack messages and clear incident notes.

Calibrate the pass bar to the band you are paying. A 20 LPA mid-level engineer who scores 70 on this rubric is a good hire. Demanding 90 at that band means you interview forever, and the market prices the 90s at senior money anyway.

Speed wins in this market. A strong Python candidate in India is in three or four processes at once, and offers land within two weeks of first contact. A five-stage loop spread over a month loses the exact people you are running it to find.

⏰ The loop that closes

Stage one, a 30-minute screen on motivation, band alignment and communication, run by us or by you. Stage two, the technical screen against the rubric from the previous chapter, 60 to 90 minutes. Stage three, your loop: one deep technical conversation with the hiring manager and one system or product conversation with a peer. Stage four, a 30-minute founder or values close. Four touches, under ten days end to end, decision inside 48 hours of the final round.

Time zones are the operational trap. India is 9.5 to 13.5 hours ahead of US time zones. The workable interview windows are early US morning, evening in India, and late US evening, morning in India. Batch your rounds inside those windows and confirm slots inside 24 hours. Every silent day between rounds reads as low interest, and candidates in this market act on that reading.

🤔 Counter-offers are the final boss

Expect the current employer to counter on resignation day, usually at 30 to 50 percent above current pay. The defence is run before the offer stage: ask directly what would keep the candidate at their current job, anchor on the reasons they engaged with you that money does not fix, and keep the gap between acceptance and joining short. Notice periods make that last part hard, which is the next chapter.

A note on how the offer itself is read. Indian candidates evaluate the whole CTC structure, not just the headline: whether PF is included, whether there is a variable component and how realistic it is, whether insurance covers parents, and whether the payslip will come from a registered Indian employer. An offer that arrives as a contractor invoice arrangement gets discounted heavily by exactly the senior people you want, because it breaks their loan eligibility, their visa paper trail and their PF continuity. Formal employment is not just your compliance posture. It is part of your offer's competitiveness.

In the US an accepted offer means a start date two weeks out. In India the standard notice period is 30 to 90 days, written into the current employer's contract, and 60 days is the practical median for mid and senior engineers. Your hiring plan needs to absorb that, because no vendor honestly makes it disappear.

⏳ The two levers that shorten it

Buyouts are the first lever. Many Indian employers allow an employee to pay salary in lieu of the unserved notice, and the new employer customarily funds it. A 60-day notice bought down to 30 costs one month of gross salary, a few thousand dollars for a mid-level engineer. Worth it when the seat is blocking a roadmap, not worth it as a default.

Pipeline timing is the second and better lever. Start the search 60 to 90 days before you need the person productive, and the notice period becomes an onboarding runway instead of a delay. Teams that plan India hiring like US hiring experience every notice period as a crisis. Teams that plan it like India hiring do not notice it at all.

One caution: a candidate who can join tomorrow is a yellow flag at mid and senior level. It usually means already resigned, benched, or between roles for a reason worth probing. The good people are employed and serving notice. The 60 days is evidence of demand, not an inefficiency.

The question every technical founder asks eventually: if my developer sits in India, do I own the code? Yes, if the paperwork is right. Indian law respects assignment of intellectual property, but the assignment has to exist, in an enforceable contract, under Indian law, with the person's actual employer.

🧾 What the employment contract must carry

Four clauses do the work. A present-tense IP assignment covering work product, hereby assigns, not agrees to assign later. A confidentiality clause that survives termination. A moral-rights waiver, because Indian copyright law grants authors moral rights that need explicit handling. And non-solicitation language that is actually enforceable, unlike broad non-competes, which Indian courts routinely refuse to enforce after employment ends. Plan your protection around confidentiality and IP clauses, not around a non-compete.

Now the structural point. A US-law agreement signed with an Indian independent contractor is a thin shield: enforcement runs through Indian courts anyway, and the contractor relationship itself is the misclassification exposure from chapter three. When the person is employed on a registered Indian entity, the IP assignment sits inside an Indian employment contract, enforceable where the person lives, with the employment relationship clean underneath it.

This is where the employer of record earns its fee on something other than payroll. Under our model the developer is employed by our Bengaluru entity with these clauses in the standard contract, and a parallel agreement assigns all work product through to you. We run this as an India-native EOR, which means the contract templates were drafted for Indian law first, not adapted from a global boilerplate after the fact.

Add the operational layer too: repos under your org, hardware policy in writing, access revocation on exit. Contracts are the backstop. Access control is the everyday protection.

Onboarding through a ready entity runs in five business days. Day one, KYC and bank verification. Day two, UAN generation for provident fund and ESI registration where applicable. Day three, equipment and access. Day four, policy induction. Day five, payroll goes live. An entity of your own takes four to six months to reach the same starting line; the sequence is documented in the hiring guide.

🧾 The monthly statutory rhythm

Indian payroll runs on a fixed compliance calendar. TDS, the withheld income tax, is due by the 7th of the following month. PF and ESI contributions are due by the 15th. Professional tax varies by state, and Karnataka expects it monthly. Quarterly TDS returns, annual PF reconciliation and Form 16 issuance stack on top. Each miss accrues interest and penalties, and each is invisible to you when the filings sit with your EOR, which is the point.

Salaries in India are paid monthly, almost always on the last working day or the 1st. Your developer will expect an itemised payslip showing gross, PF, TDS and net. Payslips matter culturally: they are the document a landlord, a bank and a visa officer all ask for. A vendor that cannot produce clean payslips is not running payroll, it is running transfers.

📅 The first-90-days management layer

The employment can be flawless and the hire still fails on integration. Set explicit written goals for days 30, 60 and 90. Assign an onboarding buddy in your home team. Overlap two to three hours of working time daily and hold one weekly synchronous one-to-one. Remote India hires fail from silence far more often than from skill.

Indian tech attrition ran 15 to 20 percent annually through the last cycle, and a strong Python engineer receives recruiter mail weekly. Losing a trained developer costs you the search, the notice period, the ramp, six months of momentum in total. Retention is not an HR nicety here. It is the difference between an India strategy that compounds and one that resets every year.

🔁 The four practices that move the number

Pay inside the band, reviewed annually against the market rather than against your home-country instincts, because a 10 percent below-market drift is a resignation letter on a delay. Give real scope: engineers who own services stay, engineers who are handed tickets leave, and this single variable outweighs the rest. Make the person visible to the wider company, in demos, in decisions, in the room where the roadmap is argued. And run the boring hygiene well, salary on the same date every month, PF filed, insurance that works, expenses reimbursed without a chase. Teams whose payroll is erratic lose people over it and never learn that was the reason.

What money cannot fix is a bad manager at headquarters who treats the India team as an execution pool. Candidates in Bengaluru talk to each other, and your company builds a reputation within two hires. The companies that win the talent market in India are the ones whose first hires tell their friends to apply.

The statutory layer helps quietly. Unbroken PF history, gratuity accruing toward the five-year mark, formal employment with a payslip trail: these anchor a developer's financial life in ways an invoice relationship never does. Properly employed people have more reasons to stay.

Plan for the exit anyway, because some attrition is structural in a market this hot. Keep bus factor above one on every service, insist on documentation as part of done, and treat a resignation as a 60-day handover project rather than a surprise, which the notice period conveniently makes possible. Teams that do this lose a person occasionally and lose momentum rarely. Teams that concentrate knowledge in one irreplaceable engineer are betting the roadmap on a coin flip every appraisal season.

The employment can be perfect and the team can still underperform if the operating rhythm is wrong. The teams that get compounding value from India engineers run three or four deliberate practices, none of them complicated.

🕐 Overlap is a design decision, not an accident

Fix a daily overlap window of two to three hours and protect it. For US East Coast teams that is early morning Eastern, evening in India. For UK teams the overlap is generous, most of the Indian afternoon. Put standups, pairing and decisions inside the window; put heads-down work outside it. Then default everything else to writing: decision logs, PR descriptions that carry context, recorded demos. A Python codebase with typed interfaces and honest docstrings is itself an async communication tool, and the same engineers who score well on the written screen from the earlier chapter are the ones who thrive in this rhythm.

🌴 Leave, holidays and the calendar you should expect

Indian employment norms give 18 to 24 days of paid leave a year plus roughly 10 public holidays that vary by state. Karnataka's list is not Maharashtra's. Diwali week is the one to plan around: assume reduced capacity, the way you would for the week between Christmas and New Year at home. Encashment of unused leave at exit is a statutory settlement item, which is one more line your payroll provider should be tracking without being asked.

🔒 Data protection is now a real obligation

India's DPDP Act puts statutory duties on companies processing personal data, and your India engineers are inside that perimeter the moment they touch production data. The practical checklist is short: access through your SSO with roles scoped to the work, production data pulled only through approved paths, company-managed devices with disk encryption, and revocation wired into the exit process. Add a written device and security policy to the employment paperwork, referenced in the contract, so the obligations are enforceable rather than aspirational. If your customers run vendor security reviews, formal employment with a registered Indian employer also answers the sub-processor questions far more cleanly than a scatter of personal-services invoices does.

None of this requires a local office. It requires deciding, once, what the rhythm is, writing it down, and holding both sides of the ocean to it. The full operating playbook, including the entity-free structure underneath it, is laid out in the no-entity hiring guide.

Everything in this guide runs through one of four structures. Here is the honest comparison, including where each one wins.

The pattern to take away: contractors win for genuinely independent work under three months. An entity wins once your India headcount is large and permanent, usually past 10 to 15 people, and we have written honestly about when that flip happens. Agencies win when you need ten temporary seats by Monday. For the team hiring its first one to fifteen full-time engineers, the EOR route is the one that is fast, clean on IP and priced flat.

One paragraph on us, and then the verdict. Versatile runs this model on our own registered Bengaluru company rather than a partner network: your developers are employed by the entity we operate, PF, ESI, professional tax and TDS filed in house across 28 states, invoices at the RBI reference rate with zero FX spread, $149 per person per month easing to $129 past twenty heads. Sourcing, when you want it, is the nine-day shortlist priced at 12 percent, invoiced at day 90. That is the whole commercial model on one line, and the full service page carries the rest.

⭐ The verdict

Hire Python talent in India for the work, not the discount. Budget the top of the band, screen on code review and communication, plan for the notice period, and put the employment on paper that survives an audit. Do those four things and India stops being an experiment and becomes the most productive line on your engineering budget.

🤔 Can I really get senior Python talent, or only juniors?

You can, if your process and band are set for it. Senior engineers with production Python, system design and clean communication exist in depth in Bengaluru, Pune and Hyderabad at 28 to 50 LPA. What they do not do is apply to vague job posts or sit in agency benches. They are reached through direct outreach and referrals, and they choose between multiple offers. If three shortlists in a row look junior, the brief or the band is the problem, not the country.

🤔 Should my first India hire be a lead or an individual contributor?

For a first hire working directly with your existing team, a strong senior IC beats a manager. You do not yet have anything to manage. From the third or fourth hire, a lead on the ground changes retention and delivery, someone who runs standups in the same time zone and gives the team a local owner. Hiring the lead first and asking them to build the team also works, but only if you have the patience for their notice period plus their hiring runway.

🤔 Do I pay Bengaluru rates if the developer lives in Indore?

Most companies land on a hybrid: one national band per level, set slightly below the Bengaluru ceiling, applied everywhere. Pure location-based pay saves money until your Indore engineer discovers the gap, and pure metro-rate pay overpays for no retention benefit. The honest answer is that talent-market pressure is national now, remote offers reach everyone, so the band that clears is closer to metro than to tier-two.

🤔 What happens if the hire does not work out?

Indian employment ends with notice, typically 30 to 60 days as written in the contract, and a final settlement covering earned salary, unused leave encashment and any gratuity due. Handled by the book it is unremarkable. Probation clauses, standard at three to six months, allow shorter notice early on, which is another reason the contract quality matters. What you avoid through proper employment is the ugly version: a disputed contractor termination with no paper trail and an aggrieved person holding your codebase access.

🤔 How many hires does it take before my own entity makes sense?

The maths usually crosses between 10 and 15 India employees, where entity running costs, 30,000 to 50,000 dollars a year in compliance, accounting and payroll overhead, drop below cumulative EOR fees. Below that line the entity is an expensive hobby. Above it, moving is sensible, and a good EOR makes the transfer boring: contracts reissued on your new entity, UAN and service dates preserved, no missed payroll in the transition. Run your own numbers in the calculator and pressure-test the answer.