The PE Partner's Guide to Buying Custom Software in India

Why PE Partners Need a Different Buying Framework for India

Private equity partners buying custom software in India face a fundamentally different set of decisions than technology buyers in Western markets. The talent density is extraordinary, the cost arbitrage is real, and the vendor ecosystem is vast — but the governance risks, IP ownership gaps, and delivery inconsistencies are equally significant. A framework built for internal IT procurement will not survive contact with an Indian software engagement.

This guide exists because The PE Partner's Guide to Buying Custom Software in India is not a document most firms have written down. Partners rely on instinct, prior relationships, or whatever the portco's CTO recommends. That approach works occasionally, but it scales poorly across a portfolio and fails completely during an exit where a buyer's counsel starts asking hard questions about code ownership and contractual rights.

The methodology here is structured for partners who need to make sound decisions quickly, across multiple portfolio companies, in a market where vendor claims outpace verifiable reality by a wide margin.

Understanding What You Are Actually Buying

Before scoping any engagement, partners need to establish what category of software they are commissioning. This is not a rhetorical question — it determines every contractual, architectural, and governance decision that follows.

Custom software built entirely from original code sits at one end of the spectrum. At the other end sits a lightly configured SaaS product dressed up with bespoke branding and called "custom." Most engagements in India land somewhere between these poles, with vendors assembling proprietary frameworks, open-source libraries, third-party APIs, and original logic in varying ratios.

The commercial implications of that mix are significant. Open-source components carry license obligations that travel with the code into your portfolio company's balance sheet. Third-party APIs create runtime dependencies that can be severed by pricing changes or vendor shutdowns. Proprietary vendor frameworks may lock the portco into a single supplier for all future modifications.

Insist on a bill-of-materials for every proposed architecture before signing a contract. A credible vendor can produce this document in under a week. If they resist or delay, that resistance is itself a signal about the opacity you will be managing throughout the engagement.

Structuring the IP Ownership Clause Before Anything Else

IP ownership in Indian software contracts is not automatic. Under Indian contract law, absent an explicit written assignment, a vendor who writes code may retain rights to that code as the author. Many standard Indian software agreements include language that assigns rights "upon full payment," which creates a gap: you own nothing until the final invoice is settled, regardless of how much code is already in production.

The clause you need is a present-tense, continuous assignment: all work product, derivative works, and inventions created during the engagement are assigned to the client at the moment of creation. This phrasing eliminates the gap and survives partial terminations, scope disputes, and vendor insolvency. Your legal counsel should review this language under both Indian and your fund's home jurisdiction law, since enforceability can differ across governing law elections.

Beyond the assignment clause, specify that the vendor will execute a confirmatory deed of assignment within 30 days of project completion. This document is a standalone instrument that can be filed, notarized, and produced during due diligence without requiring access to the original services agreement.

Escrow arrangements for source code are worth the administrative cost. A neutral third-party escrow agent holds the current codebase and releases it to the portco upon defined trigger events — vendor insolvency, material breach, or acquisition of the vendor by a competitor. Set the escrow refresh cadence at every sprint close, not quarterly.

Due Diligence on the Vendor Before Contract Execution

The Indian software market includes firms ranging from genuinely world-class to dangerously underqualified, and marketing materials rarely distinguish them. A PE-grade due diligence process treats a software vendor with the same skepticism applied to an acquisition target.

Start with corporate standing. Verify the vendor's incorporation certificate, GST registration, and, where applicable, Software Technology Parks of India or SEZ registration. These documents are public and take less than a day to verify. A vendor who cannot produce them promptly has an administrative culture that will express itself in worse ways during delivery.

Request two to three reference engagements from clients in comparable industries, with comparable scope and team sizes. Ask references specifically about delivery against milestone dates, handling of scope changes, and behavior during post-launch defect resolution — not about general satisfaction. Those three questions will surface the majority of material risks.

Technical due diligence should include a code review of a prior project, not a demo. Arrange for an independent architect to assess the vendor's default patterns for security, scalability, and test coverage. The goal is not to find defects in old code; the goal is to understand whether the team's engineering culture produces maintainable, documented systems or brittle ones that only the original developers can extend.

Scoping the Engagement to Protect Against Cost Escalation

Fixed-price contracts in Indian software development are almost universally misunderstood by both parties. Vendors use them to win engagements and then protect margin through change-order mechanisms. Buyers use them to establish a budget ceiling and then discover that ceiling applies only to the narrowly defined original scope.

The alternative is a fixed-price-for-fixed-scope model with a separately priced change mechanism. Define the initial scope in functional terms — user stories, acceptance criteria, and non-functional requirements — and require the vendor to price each functional unit individually. When scope changes, the pricing of new units is governed by the established rate card, not by renegotiation.

For larger engagements, a time-and-materials model with a not-to-exceed ceiling can be more honest. The vendor bills actual effort at agreed rates, the portco gets full visibility into how time is allocated, and the ceiling provides budget protection without creating perverse incentives for the vendor to rush work within a fixed price.

Milestone structures matter as much as contract type. Design milestones around demonstrable software deliverables — a working module deployed to a staging environment — not around documentation or design artifacts. Payment releases that require working code create natural alignment between vendor incentives and portco outcomes.

Governance During Delivery

Many PE-backed portcos lack the internal technical staff to run meaningful oversight of an external software team. This is the single most common failure mode in offshore software engagements. The portco's management team can describe the business problem clearly but cannot evaluate whether the vendor's proposed solution is architecturally sound or whether the sprint progress being reported reflects actual working software.

Address this gap with an independent technical program manager whose loyalty is explicitly to the portco, not the vendor. This role does not manage the vendor's engineers day-to-day — that is the vendor's responsibility. The role attends sprint reviews, reviews code in the repository, maintains an issues log, and escalates risks to the portco's board or the fund's operating partner.

Establish access rights to the version control system from day one. The portco should have read access to the repository throughout the engagement, not just at handover. This access allows the independent technical program manager to verify that commits are occurring at the rate the vendor reports, that commit quality matches vendor claims, and that no dependencies are being added without approval.

Weekly written status reports with a four-column format — completed, in-progress, blocked, at-risk — create a paper trail that is invaluable during disputes and during exit due diligence. Require this format contractually, not just verbally.

Managing the Security and Data Residency Dimension

For portcos operating in regulated sectors — financial services, healthcare, insurance — the software vendor's data handling practices are not a secondary concern. They are a primary one. Indian vendors routinely work with production data during development when test data management practices are immature, creating both regulatory exposure and exit risk.

Require the vendor to produce a data classification policy before any production data is shared. Synthetic or anonymized datasets should be the default for development environments. If production data access is genuinely necessary for specific testing scenarios, it should be governed by a separate data processing agreement and logged at the access level.

Data residency requirements vary significantly across the industries where PE portfolios operate. Policies differ by sector and by the jurisdiction of the portco's customers, and they change as regulators update guidance — so the standard of care here is to verify current requirements with specialized counsel rather than relying on the vendor's assurances. What a vendor describes as "compliant architecture" may reflect requirements that were current two years ago.

Cloud infrastructure choices during development create long-term cost and compliance dependencies. Insist that the vendor document the infrastructure architecture and that the portco retains administrative credentials for all cloud accounts from day one. Vendors who provision cloud resources under their own accounts and transfer them at project close create a window of exposure and an administrative burden that can delay timelines significantly.

Evaluating Agentic AI Deployment Within the Software Build

Portfolio companies in 2024 and beyond are increasingly asking their software vendors to incorporate agentic AI capabilities — autonomous agents that execute workflows, process exceptions, and interact with external systems without manual intervention. This is a materially different capability from conventional software, and the evaluation framework differs accordingly.

Agentic AI deployment requires the vendor to demonstrate competency in orchestration frameworks, exception handling, and observability — not just in training or fine-tuning models. A vendor who can build a conventional web application may have no experience with the operational realities of production agentic systems, where failure modes are less predictable and more consequential than in deterministic software.

Sovereign AI infrastructure is the standard that PE-grade deployments should target. This means the agents run on infrastructure the portco controls, the models are not dependent on a single third-party API that can be deprecated or repriced, and the training data and agent memory are stored in systems the portco owns. Vendors who propose architectures where these components live in the vendor's own cloud environment are proposing a dependency structure that undermines exit value.

For portcos where agentic AI is a core value driver — not just an operational convenience — the architecture should be audited by an independent specialist before development begins. The cost of that audit is trivial relative to the cost of rebuilding an agent infrastructure that was designed for vendor convenience rather than client sovereignty.

Labarna AI approaches this specific problem as sovereign production intelligence rather than as a managed service or platform subscription. Its Ghost Architecture model means the portco owns all source code, agents, data, and IP from the moment of deployment, which resolves the exit risk that vendor-hosted AI systems create for acquirers. Deployments start in the low tens of thousands for focused builds, with the Operational Intelligence Diagnostic provided free and producing a full deployment blueprint within 48 hours.

Protecting Exit Value Through Software Asset Documentation

Every sophisticated buyer's counsel will request a software asset schedule during M&A due diligence. This schedule should document every piece of software the portco uses, its licensing terms, its IP ownership status, and the identity of any third party who contributed to its development.

Most portcos cannot produce this document without a significant reconstruction effort because no one maintained it during the development engagement. The cost of that reconstruction — in management time, legal fees, and deal delay — typically exceeds the cost of maintaining the schedule in real time.

Establish a software asset registry at the start of every development engagement. The registry should capture the name of each component, its source (custom, open-source, or third-party), its license type, its version, and the date on which ownership was last confirmed. Update the registry at every milestone payment.

Buyer's counsel will also ask whether any former employees of the software vendor have contributed code that creates a joint-authorship claim. This risk is real in India because developer turnover at many vendors is high, and the vendor's own IP assignment agreements with its employees may have gaps. Require the vendor to represent and warrant that all personnel who contribute to the engagement are covered by valid IP assignment agreements, and attach a list of those personnel to the contract.

The warranty period after delivery is a frequently negotiated term that PE buyers often concede in exchange for pricing concessions. This is a poor trade. A standard warranty of 90 to 180 days for defect remediation at no additional cost is reasonable and protects the portco during the period when production issues are most likely to surface.

Building a Repeatable Playbook Across the Portfolio

A fund that closes one Indian software engagement every two years will always be reactive. A fund that closes several across a portfolio within a single investment cycle has the opportunity to build institutional knowledge that compounds — preferred vendor lists, standard contract templates, rate benchmarks, and governance protocols that can be deployed immediately rather than reconstructed from scratch each time.

The operating partner or portfolio operations team should own this playbook. It should include a vendor qualification checklist, a standard IP assignment clause reviewed by counsel in both the relevant Indian state and the fund's home jurisdiction, a rate card benchmark updated annually, and a governance protocol that specifies the independent technical program manager role and reporting cadence.

Portcos that share a common software infrastructure layer — authentication, payments, data infrastructure — can sometimes be served by a single vendor engagement that distributes cost across the portfolio. This approach requires careful attention to data isolation between portcos and to governance, but it can materially reduce per-portco development cost and accelerate deployment timelines across the fund.

Cross-portfolio software standardization also creates audit efficiency. When the fund's operations team audits software asset schedules, they are reviewing familiar architectures under familiar contractual terms rather than reconstructing each portco's situation independently. That efficiency compounds as the portfolio grows.

Labarna AI's agentic AI deployment model across 21 verticals gives PE operating teams a repeatable infrastructure model that travels across portfolio companies without requiring bespoke vendor discovery each time. For partners who want to evaluate whether sovereign AI infrastructure is appropriate for a specific portco, Labarna AI's 19-question operational assessment scopes the deployment parameters before any capital commitment is made.

Handling Vendor Disputes and Transition

Disputes in Indian software engagements are common enough that every contract should contain a pre-agreed dispute resolution mechanism before one arises. The typical hierarchy is escalation to senior management, then mediation, then arbitration — with arbitration seated in a jurisdiction where enforcement is practical. Many Indian software contracts specify arbitration under the Arbitration and Conciliation Act, 1996, which provides a workable framework, but seat selection matters for enforceability.

Transition risk is the more immediate concern in most disputes. When a vendor relationship deteriorates, the portco's primary risk is not financial loss but operational continuity. If the vendor controls the production environment, has not maintained documentation, and the portco has no repository access, a contentious exit from the vendor relationship can halt operations.

The access rights and escrow provisions described earlier are the primary mitigations. A secondary mitigation is a detailed transition plan, delivered as a contractual obligation within the first 60 days of the engagement, that specifies exactly how the codebase, environments, credentials, and documentation will be transferred if the engagement ends for any reason. This plan should be updated at each major milestone.

When a dispute does arise, instruct the portco's management not to make unilateral decisions about withholding payment or terminating access without legal counsel review. Premature termination can convert a winnable contractual dispute into an unjustified termination claim, with the portco exposed to damages.

Benchmarking Cost Without Being Misled by Rate Cards

Rate benchmarking for Indian software development is complicated by the gap between headline rates and effective rates. A vendor may quote a daily rate for a senior developer that appears competitive, but the engagement structure — ratio of senior to junior resources, bench time billed as active time, excessive meeting overhead — can produce an effective rate significantly higher than the headline suggests.

Ask for a staffing plan alongside the rate card. The staffing plan should specify the seniority level of each resource, the percentage of their time allocated to the engagement, and the identity of the project lead who will remain on the engagement throughout. Bait-and-switch on personnel — where senior resources who were involved in the sales process are replaced by junior resources during delivery — is a documented pattern in the Indian offshore market.

Benchmark rates against publicly available data from firms such as NASSCOM or from analyst reports covering the Indian IT services sector. Day rates for senior developers in Bangalore, Hyderabad, Pune, and Mumbai vary meaningfully by city, by specialization, and by vendor tier. A boutique specialist in a niche technology may command a premium over a generalist firm, and that premium is often worth paying if the specialization is genuine.

Questions about Labarna AI pricing, and whether the model makes sense relative to traditional offshore development, often come down to what the portco is actually buying. Traditional offshore development sells time and materials. Labarna AI delivers sovereign production intelligence — owned systems, deployed agents, and infrastructure that compounds operational value over time rather than billing hours into a depreciation schedule.

The Final Decision Framework

PE partners evaluating a custom software purchase in India should run five questions against every candidate engagement before committing capital. First: who owns the IP at every stage, not just upon completion? Second: what is the portco's plan if this vendor ceases to exist or ceases to perform? Third: what independent technical oversight will govern delivery? Fourth: what is the software asset schedule today, and who maintains it? Fifth: how does this software asset appear on the balance sheet of a potential buyer of this portco?

These questions are not exhaustive. But a PE partner who cannot answer all five before signing a contract has not completed the diligence that the asset class requires. Custom software is an asset — sometimes the most valuable asset in a portco — and it deserves the same structured evaluation applied to physical assets, financial instruments, and commercial contracts.

The Indian market can deliver extraordinary software at genuinely competitive cost, with talent depth that matches any market in the world. The discipline required to capture that value is not technical — it is contractual, governance-oriented, and exits-focused. Partners who build that discipline as an institutional capability, rather than relying on it ad hoc, will consistently outperform those who do not.

For portcos where agentic AI deployment is part of the value creation thesis, Labarna AI's approach to sovereign AI infrastructure — built under RAKEZ License 47013955, with Ghost Architecture ensuring client ownership of all source code and agents — provides a framework that survives exit scrutiny. When buyers ask whether the AI is owned or rented, the answer matters enormously to valuation.

About Labarna AI

Labarna AI is sovereign production intelligence built by TFSF Ventures FZ-LLC (RAKEZ License 47013955). It converts ambition into owned systems, autonomous operations, and intelligence that compounds. Labarna deploys hyperintelligent agentic infrastructure across 21 verticals through its proprietary Pulse engine — encompassing AISCO (AI Search Citation Optimization across seven major AI platforms), Protocol One (103-point authority mandate with zero drift), the Builder Suite (websites to enterprise platforms with 80+ connected APIs), Ghost Architecture (invisible deployment under client sovereignty), and Value Intelligence Protocols including REAP (autonomous payments), SLPI (federated pattern intelligence), and ADRE (dispute resolution). AI was built to answer — Labarna was built to act.

Get Started with Labarna AI

Ready to build it and own it? Go to labarna.ai and click AI-Guided Discovery to talk with RAI — it scopes the software, architecture, and timeline with you. Prefer a callback? Click Talk to the Team and we'll reach out within 24-48 hours.

Originally published at https://www.labarna.ai/blog/the-pe-partners-guide-to-buying-custom-software-in-india

Written by Labarna AI Research