Layer 1: Physical Office Infrastructure

The physical environment of an ODC is not just workspace — it is the operational foundation on which every other infrastructure layer is built. Getting the physical setup wrong creates cascading problems for IT, network, and security infrastructure.

Commercial space selection for an ODC

ODC space requirements differ from general commercial office requirements in several critical ways:

• Power load capacity: An engineering team’s IT equipment consumes significantly more power per workstation than a standard office. Server racks, high-performance workstations (common in development environments), UPS systems, and cooling infrastructure for the server room require an electrical load analysis before signing a lease. Many commercial buildings in India have insufficient load capacity for dense IT environments without expensive upgrades.
• Server room viability: The space must have — or be capable of having — a dedicated server room (or server alcove for smaller setups) with adequate ventilation, cooling, and secure access. Not every commercial floor plate accommodates this without major structural intervention.
• Structural cabling capacity: Raised floors or sufficient ceiling pathways for extensive LAN cabling are a practical consideration for larger ODCs. Buildings with limited cable management infrastructure create installation complexity and ongoing maintenance issues.
• Lease flexibility for growth: An ODC that starts at 20 seats and grows to 80 over two years needs either adjacency options (available expansion space on the same floor or building) or a lease structure that allows relocation without penalty. Locking into a space with no headroom is one of the most expensive infrastructure decisions a company can make.

Workstation and ergonomic fit-out

Engineering teams spend 8–10 hours daily at their workstations. Ergonomic quality — chair specification, monitor height, keyboard and mouse setup, desk surface area — directly affects productivity and health outcomes over a multi-year engagement. This is not a cost-cutting area.

For a software development ODC, workstation fit-out typically includes:

• Sit-stand or height-adjustable desks (increasingly standard for engineering environments)
• Multi-monitor arm mounts (most developers work with dual monitors; some roles require three)
• Quality ergonomic chairs with lumbar support rated for extended use
• Acoustic panels or partitions between workstation clusters to reduce noise in open-plan environments
• Dedicated collaboration zones distinct from focus work areas

Conference and collaboration infrastructure

ODC teams operate in a hybrid collaboration model — local team in India working synchronously with the parent company team in another timezone. Conference rooms must be equipped for high-quality video conferencing, not the basic webcam-and-laptop setup that is standard in serviced offices. This means:

• Ceiling or bar-mount microphones that capture room audio clearly
• Wide-angle cameras capable of showing all room participants
• Displays sized appropriately for the room (a 65-inch display in a 10-person conference room vs. a 32-inch monitor in a 4-person huddle room)
• Acoustic treatment that prevents echo and ambient noise from degrading call quality

These are not peripheral considerations. For a team whose primary collaboration channel is video, poor audio-visual infrastructure is a daily operational friction that compounds over months.

Layer 2: IT Infrastructure

IT infrastructure for an ODC must be specified before the physical fit-out begins — not after. This sequencing rule is violated constantly, and it is one of the primary causes of ODC launch delays.

The reason is dependency sequencing: structured LAN cabling must be planned alongside the partition layout. Server room location and power requirements must be confirmed before electrical load planning is finalized. Endpoint specifications must be decided before procurement begins, since high-specification developer laptops have lead times.

Server room or server rack setup

Whether the ODC maintains on-premises servers or operates entirely in the cloud, a dedicated server area is necessary. Even cloud-first environments need on-premises networking equipment — switches, firewalls, UPS units, and patch panels — that require a controlled environment.

For ODCs with on-premises build servers, test environments, or local storage:

• Rack-mounted server infrastructure sized for current load plus 50% growth headroom
• Dedicated UPS with adequate runtime for graceful shutdown during power interruptions
• In-row or room-level cooling adequate for the rack thermal load
• Physical access control to the server room (biometric or keycard, separate from main office access)
• Environmental monitoring — temperature and humidity alerts

For cloud-based ODCs, the server room simplifies to a managed network rack, but the requirement for controlled environment, adequate power, and physical security remains.

Structured LAN cabling and network infrastructure

Wireless connectivity is sufficient for general office use. It is not sufficient as the sole connectivity medium for a software development environment. Engineers running large repository pulls, deploying to staging environments, or streaming high-quality video in collaboration sessions require wired network connectivity with consistent performance.

Structured cabling specification for an ODC should include:

• CAT6A or CAT6 cabling to each workstation point (CAT5e is not recommended for new installations)
• Network switch placement that minimizes cable run lengths and allows efficient future expansion
• Patch panel organization that makes moves and changes manageable without disrupting active network segments
• Separate VLANs for workstations, server infrastructure, guest access, and IoT/management devices — network segmentation is a security requirement, not just good practice

Endpoint provisioning

Endpoints — laptops, desktops, monitors, and peripherals — should be specified by the parent company’s IT team and procured before hire start dates are confirmed. Common mistakes:

• Procuring endpoints after engineers join, which means engineers wait days or weeks for equipment
• Allowing engineers to use personal devices during the waiting period — a serious endpoint security gap
• Specifying standard business-class hardware when the team is running memory-intensive or compute-intensive development workloads

High-performance developer workstations in India cost significantly less than equivalent hardware in Western markets, making this an area where the ODC model delivers cost efficiency without compromising tooling quality.

Layer 3: Network and Connectivity Infrastructure

Network infrastructure is the highest-stakes, longest-lead-time component of ODC setup. It is also the component most often treated as something to handle “after the office is ready.”

Internet leased line provisioning

An internet leased line (ILL) — a dedicated, symmetric, uncontended connection from an ISP — is the standard connectivity choice for an ODC. Unlike consumer or business broadband, an ILL guarantees consistent upload and download speeds, which is essential for a team pushing code, accessing cloud environments, and running video collaboration.

Key considerations:

• Lead time: ILL provisioning in India typically takes 15–30 working days after order confirmation. This is not negotiable — it is an ISP operational reality. Orders must be placed before or immediately after lease signing, not after fit-out is complete.
• Bandwidth sizing: 10–20 Mbps per concurrent user is a reasonable baseline for a software development environment. For teams with heavy cloud usage, video conferencing, or CI/CD pipelines, 30+ Mbps per user is more appropriate. Undersizing bandwidth is a productivity constraint that is expensive to fix quickly — ILL upgrades take time.
• Redundancy: A primary ILL plus a secondary broadband or separate ILL link on a different physical infrastructure path. Single-point connectivity is an unacceptable business continuity risk for a development centre.
• SLA: Enterprise-grade ILL contracts include uptime SLAs (typically 99.5%–99.9%), support response time commitments, and escalation paths. Consumer-grade broadband does not.

VPN configuration and secure connectivity

An ODC’s engineers must connect to parent company systems — version control, project management tools, internal APIs, staging environments — through a secure, encrypted channel. VPN configuration connects the ODC to headquarters systems in a way that is both secure and performant.

Common VPN architectures for ODCs:

• Site-to-site VPN: A persistent encrypted tunnel between the ODC’s firewall and the headquarters network. All traffic between sites travels through this tunnel. Best for teams with frequent access to on-premises parent company systems.
• Client VPN with split tunneling: Individual engineers connect to headquarters VPN for access to restricted resources while routing general internet traffic directly. Better for cloud-native environments where most tools are SaaS.
• Zero Trust Network Access (ZTNA): Increasingly adopted for distributed engineering environments. Each access request is authenticated and authorized individually rather than relying on network perimeter trust. More secure than traditional VPN architectures for environments handling sensitive data.

The choice of architecture should be made by the parent company’s IT security team — not the local setup vendor. The local infrastructure must be designed to support whichever architecture is chosen.

Firewall and network security

An enterprise firewall/UTM (Unified Threat Management) appliance provides application-level traffic inspection, intrusion detection, web content filtering, and network segmentation enforcement. This is standard equipment for any ODC handling client data, proprietary code, or regulated information.

Layer 4: Physical and Data Security Infrastructure

Security infrastructure for an ODC serves two distinct purposes that are often conflated: physical security (controlling who can access the facility and monitoring what happens inside) and data security (controlling what data can leave the facility and in what form).

Both are non-negotiable for a software development centre.

Physical security

• Biometric access control: Fingerprint or facial recognition-based access at the main office entry and at the server room. Card-based systems are acceptable but biometric is more appropriate for environments handling sensitive IP. Access logs must be retained and reviewable.
• CCTV surveillance: Full coverage of workstation areas, server room entrance, main entry and exit, and emergency exits. Footage retention of at least 30 days. Remote access to live feed for management review (with appropriate privacy notice to employees).
• Visitor management: A formal visitor registration process — not a paper sign-in sheet — with identity verification and host notification. Temporary access badges that are clearly distinguishable from employee access credentials.

Data security infrastructure

• Endpoint encryption: Full-disk encryption on all endpoints using BitLocker (Windows) or FileVault (macOS). Required before any production data or code is accessed on the endpoint.
• Mobile Device Management (MDM): Enrollment of all endpoints in an MDM platform (Microsoft Intune, Jamf, or equivalent) allows remote wipe in the event of loss or theft, policy enforcement, and software inventory management.
• Data Loss Prevention (DLP): For ODCs handling sensitive client data or regulated information, DLP tools monitor and block unauthorized transfer of sensitive data to external storage, personal email, or cloud storage.
• USB and peripheral control: Blocking unauthorized USB storage devices is a standard endpoint security control. This is typically enforced through MDM policy or endpoint security software.

Layer 5: Operational Infrastructure

Physical, IT, network, and security infrastructure creates a workspace. Operational infrastructure makes it function as a business.

Power backup — UPS and DG

Power reliability varies across commercial buildings in India. Even in well-serviced micromarkets in major cities, brief power interruptions occur. For a development centre, an unplanned power cut that takes down servers and endpoints mid-development session is not just inconvenient — it can corrupt active work, interrupt CI/CD pipeline runs, and create data consistency issues in shared environments.

• UPS (Uninterruptible Power Supply): Provides instantaneous power bridging during utility power interruptions. Sizing must be based on actual IT load (server, networking equipment, and workstations), not estimated load. UPS batteries must be tested and replaced on schedule.
• Diesel Generator (DG): For longer power interruptions, a DG set provides sustained backup power. Most Grade A commercial buildings in India include shared DG backup, but the available load must be confirmed before signing the lease. Some buildings allocate DG load on a shared basis that may be insufficient during peak outage periods.