Challenge
The leadership team had decided India was the right long-term base for a second engineering hub, but the direct-captive route looked daunting: registering an entity, statutory compliance, real-estate and a 12-to-18-month setup before meaningful output. Doing it wrong risked permanent-establishment and tax exposure they were not equipped to manage from Dublin. They also did not want a classic outsourcing arrangement where the talent, the processes and the institutional knowledge would never actually be theirs. What they needed was a way to ramp a real team quickly now, with a clean, contractual path to owning it later.
OSCABE approach
OSCABE ran a Build-Operate-Transfer engagement: we built and operated an 18-person captive in Hyderabad on the client's behalf, with OSCABE acting as employer of record so they avoided standing up an entity on day one. Every hire, from platform and backend engineers to a QA function and an SRE, went through our 5-stage vetting and reported into the client's own engineering managers with a 4 to 6 hour daily overlap. We handled payroll, statutory compliance, facilities and HR through the operate phase, with a pre-agreed transfer clause covering people, IP, tooling and processes. The whole thing ran under one UK contract until the client was ready to take ownership.
Outcome
The first cohort was live and shipping inside about sixteen weeks, against the year-plus a direct entity would have taken. Fully loaded run-cost per engineer landed roughly 60% below the equivalent Dublin seat, and the client retained full IP and the option to transfer the team into their own Indian entity on a defined timeline of around 18 to 24 months. Retention through the build and operate phases has held well, and the captive now owns two product lines end to end. Leadership describes the model as getting a captive's control without a captive's upfront risk.
Inside the engagement
The full evidence: team, timeline, stack, vetting, security, costs and before/after metrics.
The problem
The client, a Dublin-headquartered B2B SaaS scale-up, had decided India was the right long-term base for a second engineering hub but had no appetite for the direct-captive route: registering an entity, statutory compliance, real estate and a 12-to-18-month setup before meaningful output, with permanent-establishment and tax exposure they were not equipped to manage from Dublin. Equally, they did not want a classic outsourcing arrangement where the people, processes and institutional knowledge would never become theirs. They needed to ramp a real team quickly now, with a clean contractual path to owning it later.
Team composition
OSCABE built and operated an 18-person captive in Hyderabad, structured to run two product lines:
- 1 engineering manager (13 yrs, owned local delivery and reported into Dublin)
- 6 backend engineers (mix of 5 to 10 yrs, Java and Go services)
- 4 frontend engineers (4 to 8 yrs, React and TypeScript)
- 2 SRE / platform engineers (8 and 6 yrs, Kubernetes and observability)
- 3 QA engineers (one lead at 7 yrs plus two automation engineers)
- 1 product analyst and 1 scrum master
All reported into the client's own engineering managers on a 4 to 6 hour daily overlap with Dublin.
Timeline
The first cohort was live and shipping inside sixteen weeks:
- Weeks 1 to 4 (Build): site, payroll and statutory setup via OSCABE as employer of record; first manager and senior engineers hired and vetted.
- Weeks 5 to 10 (Build to Operate): remaining roles filled; onboarding onto the client's codebase, CI and runbooks; first product line picked up.
- Weeks 11 to 16 (Operate): full team shipping against the Dublin roadmap, owning a second product line, with retention and HR processes embedded.
- A pre-agreed transfer window of around 18 to 24 months governs the eventual move into the client's own Indian entity.
Tech stack
Java (Spring Boot) and Go microservices on PostgreSQL and Kafka; React with TypeScript front ends; Kubernetes on AWS (EKS) with Terraform infrastructure-as-code; GitLab CI for build and deploy; Prometheus and Grafana for observability. The captive worked the same stack as Dublin so code and ownership moved cleanly.
How OSCABE vetted the team
Every one of the eighteen hires cleared OSCABE's five-step vetting. An instant AI shortlist matched candidates to each role brief; a senior OSCABE specialist ran a technical interview with a live coding or system-design exercise on the client's stack; a communication assessment confirmed they could operate inside a Dublin-aligned, English-language engineering org; background and reference checks validated history; and Chartered Engineer / CE verification was applied to the engineering roles. The engineering manager was additionally assessed on people-leadership and delivery track record.
What was delivered
A fully operational captive engineering centre shipping production software inside four months, against the year-plus a direct entity would have taken. The captive now owns two product lines end to end, including their roadmaps, on-call and quality. Just as important, the client retained full IP and a defined contractual path to transferring the people, tooling and processes into their own Indian entity on their timeline.
Client workflow and collaboration
The captive operated as a second engineering hub, not a vendor. Engineers sat in the client's sprints, code review and on-call rotations, with the Hyderabad engineering manager reporting into the Dublin VP of Engineering. Quarterly business reviews tracked delivery, retention and the transfer roadmap so leadership always knew exactly what they would be taking ownership of.
Tools used
Jira and Confluence; GitLab for source, review and CI; Slack and Google Meet for comms; PagerDuty for on-call; Lattice-style performance tooling run by OSCABE during the operate phase; and a shared transfer-readiness tracker covering people, IP, contracts and tooling.
Security and compliance model
The engagement ran under ISO 27001-aligned controls and EU GDPR, under one UK contract through the operate phase. Engineers used SSO with least-privilege access on managed, encrypted devices. Cross-border data flows between the EU and India were covered by SCCs plus a transfer risk assessment. OSCABE acted as employer of record, carrying Indian statutory compliance, payroll and HR, which removed the client's permanent-establishment risk until transfer. Every employee signed an NDA and an IP-assignment clause running to the client, and the transfer agreement carried that IP cleanly into the client's entity.
Cost comparison
| Item | Dublin in-house (per FTE) | Hyderabad captive (per FTE) |
|---|---|---|
| Fully loaded annual run-cost | around EUR 110,000 | around EUR 44,000 |
| 18-FTE annual run | around EUR 1,980,000 | around EUR 790,000 |
| Reduction per FTE | about -60% |
Before and after
| Metric | Before | After |
|---|---|---|
| Time to first shipping cohort | 12+ months (direct entity) | 16 weeks |
| Run-cost per FTE | around EUR 110,000 | around EUR 44,000 |
| Entity / PE risk on day one | high | carried by OSCABE |
| IP ownership | n/a | retained by client |
| Product lines owned by captive | 0 | 2 |
What OSCABE managed vs what the client managed
OSCABE managed:
- Site setup, employer-of-record status, Indian payroll, statutory compliance, facilities and HR.
- Hiring, five-step vetting and retention across all eighteen roles during build and operate.
- The security baseline and the transfer-readiness process.
The client managed:
- Engineering direction, roadmap and acceptance of all work through their own managers.
- Architecture, code review standards and on-call ownership.
- The decision to trigger transfer and the receiving entity.
Why remote worked
A captive only works if the team is genuinely the client's, and that is about reporting lines and ownership, not geography. The Hyderabad team sat in Dublin's sprints and on-call on a 4 to 6 hour overlap, ran the client's exact stack, and built two product lines they fully own. OSCABE carrying the entity, payroll and compliance through the operate phase meant the client got a captive's control and IP without a year of upfront entity risk, with a clean path to taking it in-house when ready.
“We wanted our own team in India, not a vendor relationship, but we could not justify a year of entity setup before writing any code. OSCABE built and ran the centre for us, and we keep the people and the IP at the end. It is the captive we wanted without the upfront risk.”