• Blog

    • Onshore VS Nearshore: talent location evolution for AI augmented programs

    Sedona Digital

    June 19, 2026

    This article is written and published in conjunction with Sedona Digital

    Onshore vs Nearshore?

    The Five key Influencing Trends

    • Business SMEs and leadership are often office-based or hybrid workers near your UK sites - they're critical for approvals, process augmentation (BA) and application design, KPI setting and rapid feedback/testing when you're straight into UX or improving core business KPIs.

    • AI application, data and cloud architectures are evolving rapidly with distributed tech leadership, service and product partners all opinionated on priorities, and often remote and misaligned. When your AI and supporting data frameworks are new or evolving quickly, the architects and lead engineers are in daily design, change and approvals management as well as supporting POCs and vendor product analyses.

    • Mixed discipline scrum teams are still redefining roles and responsibilities with AI-assisted backlog, dev and QA, and are often located remotely from architecture and business SMEs. In most cases, culture or skills changes need planning with line management and HR as much as technical steering.

    • Growing numbers of agentic coding 'orchestrators' are starting to cause talent drain of traditional engineers with tacit domain knowledge, increasing the need for higher-affinity learning between new agentic orchestration coding devs and business SMEs, with AI-savvy POs.

    • Higher-paced development is overtaking requirements quality and management as the old fortnightly backlog scrum with POs is too late for the code agents. Where bypassing this is prohibited, requirements management is now the new bottleneck,not coding as agentic maturity in backlog curation lags gen, vibe or low coding.

    So Where Should You Place Your Team? 

    So how does this determine where you may want to place your team? Here's a few examples considerations that reflect current trends

    Designer-1

    BAs & POs

    Whether it's mapping out complex process augmentation and optimisation, wireframing the right level of agentic observability for user trust and adoption, challenging meaningful KPI value, or having the hard conversations about what goes into this week's backlog and what doesn't, there have never been more reasons for a BA or PO to work face-to-face with your business SMEs. Onshore makes this easier and increases the likelihood of national sector experience and cultural alignment. However, if you're only looking to start with one or two simple applications or use cases a quarter, and your BAs or POs are happy to travel abroad every week or two, then saving a few hundred euros a day may be worth it. It is important, however, to consider rising travel and expenses costs, the risk of tiring them out, and the value of retaining knowledge close to the business as demand for AI applications inevitably grows. 
    Designer-1
    Designer (1)

    AI Tech Architects

    There's a reasonable chance that your architectural frameworks are undergoing a lot of change and a surge towards maturity. New agentic API management options at the cloud layer change as fast as your catalogue of LLM services and respective models. Your data framework now needs to address not just medallion, chunking, and pipelines, but also advanced metadata for agent relevance, token cost control, and hallucination reduction. RAG architecture and grounding data are being pushed in by business SMEs. The amount of observability data coming from different parts of the architecture is growing rapidly. Classification for sensitivity, eligibility, and relevance may also need an overhaul, now that the agents can make use of so much data. SOA and microservice practices have never been needed more, yet the coding agents are producing monoliths with increasing token development usage and blurred interoperability. Lead AI Architects are becoming masters of not just the latest technology options from the big vendors, but the qualitative aspects of managed data framework nodes that strong AI workflows in production demand. Regular travel to whiteboard, collaborate, and achieve consensus is becoming increasingly common, with a growing number of stakeholders. That said, tenacious nearshore architecture leads can work well too, as breaking out of the 'London-centric' beat can help reach out to collaborate and agree with new, more distributed parties. 

    Designer (1)

    QA & Data Engineering

    QA & Data practices are evolving globally, with traditional diligence modernising to make metadata engineering a first-class citizen, and agentic QA an integral part of agentic development. Both may leverage data platforms for model drift management against 'gold set' results, prompt versioning and retention, RAG and observability frameworks with reference and grounding data. Internal QA data consumers are expanding from IT to compliance, risk, and audit. High-quality, process-integrated talent can be found nearshore, and they collaborate very productively around existing SDLC processes and your AI data framework. Modern engineering quality practice and horsepower benefit hugely from having multiple people in the same practice in one location, sharing knowledge and intellectual property across totally different programmes. For this reason, consuming a trusted partner's nearshore QA and data engineering practice makes great sense, and good partners go the extra mile to adopt your tools, culture, and practices quickly. The 30–40% cost saving versus UK onshore becomes the icing on the cake, rather than the only benefit sought. 

    Stability and Talent Retention

    Another key consideration is contract vs permanent, or long-term/dedicated, and how your preference relates to regional market trends and talent availability. Whilst AI agents operating at the business process level may reduce some need for domain process expertise, your industry, workflow, tools, and organisational knowledge is still incredibly valuable in many roles, on top of technology experience. It's likely much of this will be learned by new people on your time, so we want to keep the people we attract with low churn, certainly for the first 24 months, to see a return on your onboarding overheads. We see that, as well as culture, talent, and experience, certain markets have a higher preference for permanent versus contract, to give you that stability as a direct employer or when consuming tenured vendor services. At Sedona, we've found that currently Eastern Europe is proving to be a great area for us to employ AI engineering talent, for the sort of roles that work nearshore for our customer teams. The combination of quality, availability around Microsoft Azure technologies, a preference for permanent versus contract, and a cultural fit with Financial Services organisations is hard to beat in Romania, Serbia, and similar markets. Our nearshore teams also work seamlessly with our onshore resources – those BAs, EAs, and PM/POs who are on customer sites every day. 

    We're able to learn from a myriad of different AI projects for different customers, as well as from managing different talent pipelines all over the world. Our customers demand quality, cultural alignment, rapid integration, and value. If you want to discuss nearshore vs onshore for Financial Services AI, feel free to get into contact, as we have the experience to help you understand your opportunities and avoid the pitfalls.

     

    Get in touch to find out more