Engineering services contracts rarely make headlines in the network equipment world. They are usually treated as operational housekeeping: capacity added here, a support programme moved there. The five-year agreement between Nokia and L&T Technology Services (LTTS) deserves a closer reading, because its scope says a great deal about where telecom R&D is heading.
Under the agreement, LTTS will set up a dedicated offshore development centre in India and build an engineering team working alongside Nokia‘s global R&D organisation. The remit spans product engineering, new product development, embedded systems, software development, system integration and ongoing product support. The detail that stands out is the centre’s technical focus: combining data-centre-grade compute with AI infrastructure engineering for next-generation telecom solutions, across 5G networks, intelligent automation and AI-driven operations. Financial terms have not been disclosed.
Read alongside Nokia’s recent performance, the deal is a useful window into how telecom R&D is being rewired for an AI-shaped market.
Telecom R&D now has a data centre dimension
For most of the 5G era, vendor R&D was organised around the radio access network, the packet core and the transport layers that connect them. That work remains essential, but the centre of gravity is shifting. Nokia’s Q3 2026 interim report shows AI and Cloud customer net sales of EUR 631 million, up 126% year on year and now roughly 12% of group sales. Order intake from the same customer group more than doubled. Optical Networks grew 18% at constant currency and IP Networks grew 43%, and the company expects its combined optical and IP businesses to grow by 25% to 30% in 2027.
That growth changes what engineering teams are asked to build. Hyperscale and AI data centre customers care about throughput per watt, interconnect density, automation at scale and integration with compute platforms. Telecom operators increasingly want the same qualities in their own networks as AI inference moves closer to the edge and network operations become more automated. An engineering centre that explicitly blends data-centre-grade compute with telecom product development reflects that convergence. The skills that design a coherent optical module, an AI-assisted network management system or an accelerated RAN platform are drawing from the same pool.
How telecom R&D operating models are changing
Nokia’s R&D investment is substantial, with reported R&D expenses of more than EUR 1.2 billion in the third quarter alone. At that scale, the question is less about how much to spend and more about how engineering capacity is structured, located and directed. Nokia has also said it has increased the scale of programmes to simplify its operations and location strategy, alongside the integration of Infinera, which is expected to complete by the end of 2026.
An integrated partner model offers several advantages in that context. A dedicated centre can be ramped up or redirected as priorities move between product lines, without the lead times of building internal teams from scratch in each discipline. Partner engineers working inside the vendor’s own processes keep the knowledge close to the product, which distinguishes this model from older outsourcing arrangements built around discrete, ticketed work packages. A five-year term provides enough continuity for the partner team to accumulate deep product familiarity, which is particularly valuable for embedded systems and long-lifecycle network platforms that need years of sustained support.
The breadth of scope is telling too. Covering everything from new product development to lifecycle support means the centre is positioned as a full engineering partner across the product journey, rather than a single-function resource.
Where AI-era telecom R&D demands new skills
The shift is easiest to see when traditional telecom engineering priorities are set against what AI infrastructure now requires:
| Engineering domain | Traditional telecom R&D focus | AI-era requirement |
| Compute platforms | Purpose-built baseband and network appliances | Data-centre-grade compute, accelerators and cloud-native platforms |
| Optical and interconnect | Metro and long-haul transport capacity | High-density data centre interconnect, pluggable coherent optics, power efficiency |
| Network operations | Rule-based management and manual workflows | AI-driven operations and intelligent automation at scale |
| Embedded systems | Radio and transport hardware firmware | Firmware for heterogeneous, accelerator-rich hardware |
| Software and integration | Vendor-specific network functions | Open, API-driven integration across telecom and cloud stacks |
Each of these areas is in strong demand across telecom, cloud and semiconductor companies simultaneously. Vendors that can secure dedicated, scalable access to that talent through long-term partnerships are better placed to move quickly as product roadmaps evolve.
India’s role in global telecom RD
India has long been one of the most important engineering bases for network equipment vendors, and Nokia’s presence there spans R&D in Bengaluru alongside major operations in Chennai, Gurugram, Mumbai and Noida. Nokia’s engineering leadership has described the country as an important innovation hub and framed the LTTS partnership as a way to expand its R&D footprint and accelerate development.
What is changing is the type of work India-based teams are taking on. The country’s engineering ecosystem is moving up the value chain from sustaining engineering and testing towards product ownership in advanced domains, including AI infrastructure, semiconductors and cloud-native software. A centre explicitly dedicated to next-generation compute and AI infrastructure for telecom fits that trajectory, and it places India-based engineers directly in the fastest-growing part of Nokia’s business.
What the deal means for engineering services in telecom R&D
For LTTS, the agreement aligns with its Lakshya 31 strategy, which identifies next-generation compute and AI infrastructure as one of its Big Bets. The company brings considerable scale to that ambition, with more than 24,000 employees, 21 global design centres and 111 innovation labs, and a client base that includes 69 Fortune 500 companies and 57 of the world’s top ER&D spenders. LTTS described the engagement as a significant milestone for its Technology segment.
The broader pattern is that engineering research and development providers are repositioning from capacity suppliers to domain specialists. The firms winning multi-year mandates from network vendors are those that can demonstrate depth in the specific technologies where their clients are investing, rather than general engineering bandwidth. Expect more agreements framed around AI infrastructure, optical systems and network automation, and fewer framed around generic software services.
Signals to watch in telecom R&D partnerships
Several developments will indicate how far this model extends across the industry:
- Vendor adoption. Whether other network equipment vendors announce similar long-term, domain-specific engineering partnerships tied to AI and data centre products.
- Depth of responsibility. How quickly partner-led centres move from development support into ownership of complete product modules or platforms.
- Scope of mandates. The share of telecom engineering mandates that explicitly reference data centre, optical or AI infrastructure scope.
- Customer mix. Whether India-based telecom R&D increasingly feeds products sold to hyperscale and AI customers, not only to operators.
- AI in engineering itself. How AI-assisted engineering tools change productivity inside these centres, and how that reshapes partnership structures over a five-year term.
The TeckNexus view on telecom R&D
The Nokia and LTTS agreement is a modest announcement on its own, but it captures a structural shift. Telecom R&D is no longer a closed discipline focused solely on operator networks. It is becoming part of a wider AI infrastructure engineering effort, where optical, IP, compute and automation expertise overlap, and where vendors build capacity through deep, long-term partnerships as well as internal teams. The vendors and engineering partners that organise around that convergence early will shape how the next generation of networks, inside and outside the data centre, is designed and built.
For more analysis on 5G, AI infrastructure and advanced connectivity, subscribe to TeckNexus Monthly Insights. If you would like to share your perspective with our audience and generate qualified leads, partner with TeckNexus through our Intelligence Programs.
