Building an AI-Ready Organisation: Lessons from Sigma Software Group

Most companies now claim to adopt AI, yet few can demonstrate tangible business results. According to McKinsey & Company, 88% of organisations globally now report using AI in at least one business function, and 79% regularly use generative AI (GenAI) in at least one function. In our experience working with enterprises and scaling AI inside our own organization, the gap between expectations and outcomes is far wider than public reports suggest. Real progress requires more than access to tools, it mandates operational change, disciplined experimentation, and a clear understanding of where AI can create measurable value. 

Over the past decade, our experience at Sigma Software Group has demonstrated that building an AI-ready organisation is less about the technology itself and more about structure, culture, and timing. In other words, AI adoption is not a system integration, it’s a transformational project.  Our experience highlights four factors that consistently matter.

1. Start with people, not tools

The first step was not to automate everything overnight but to help employees understand what AI could already do. Training people to use tools like ChatGPT or Microsoft Copilot created a shared foundation. It allowed everyone — from HR to marketing and finance — to learn capabilities and limitations of the tools and see AI as a practical assistant rather than an abstract concept.

To speed up adoption in our organisation, we launched the AI Champions program. We selected proactive employees from each unit, gave them the authority and time to implement changes, and started with a three-day bootcamp focused on showcasing best use cases, learning prompt engineering and how to build agents. We also provided advanced tools so they could run experiments on their own within their units.

Using AI tools in everyday tasks is a new mandatory skill for all our employees. By becoming proficient in this area, we were able to move forward with agentic automation.

2. Apply AI in core processes

As a software engineering company, we began exploring GenAI in our core processes, i.e. the software development lifecycle. We piloted tools like GitHub Copilot and Cursor across a small set of projects, established clear metrics to measure their impact, and supported engineers through workshops and a growing knowledge base of best practices.

Achieving reliable results aligned with global findings allowed us to scale this practice across all our projects. Our internal research shows AI code generation tools helped our engineers save 16.8% of their time, with 27.5% of AI-generated code accepted as is.

We also identified limitations that prompted us to develop additional tools for the software development lifecycle, including autonomous unit-test generation, code change impact assessment for regression testing, and an AI project-team assistant designed to support onboarding and help teams navigate shared project context.

Continuous learning, trend monitoring, knowledge sharing, and targeted R&D efforts now underpin our improvement of core processes.

3. Share experience

Through close collaboration with our clients and partners, we identify common areas for the best AI applications. Our experience has shown that many organizations encounter similar challenges during their AI adoption journeys.

According to MIT study, only about 5% of organizations see meaningful returns from GenAI adoption. Of those, two-thirds succeed through strategic partnerships with technology firms. This is unsurprising, given how new and complex GenAI is. The organizations that move fastest are the ones that work with partners who have already delivered similar projects instead of trying to figure everything out internally.

We have been open to sharing our learnings and insights from our own AI transformation with our clients. This support takes various forms, including joint training sessions and workshops, assistance in developing AI strategies and roadmaps, as well as opportunities to participate in our ongoing R&D initiatives and pilot projects.

This collaborative approach leverages our technical expertise with clients’ industry insights to deliver AI-driven solutions for sectors including banking, insurance, telecom, ad tech, and health tech.

One such example is a document processing solution that we developed together with our client, and now use this approach as a backbone for similar tasks, helping save time and money in future projects.

4. Run controlled experiments

A culture of continuous experimentation is essential for AI adoption, but without structure, it quickly becomes counterproductive. Many organisations end up running dozens of disconnected proof-of-concepts (PoCs) with overlapping goals, incompatible technologies, and no clear path to implementation.

Effective experimentation requires governance. Each PoC must have a defined objective and, ideally, a business case. Organisations also benefit from structured practices that channel creativity without losing control, such as regular hackathons, ideation workshops, and internal showcases to share results, prevent duplication, and enable teams to build on one another’s work.

In addition to managing PoCs, organizations need to develop expertise and create solutions. Therefore, we also strategically allocate dedicated R&D resources to develop solutions that with clear, demonstrable value in these scenarios both internally as well as for our clients. This disciplined experimentation framework has produced tools for automated unit-test generation, code-change impact analysis, onboarding assistants. One of these initiatives, our corporate AI assistant SIMA, has even successfully matured into a marketable product.

Ultimately, what matters the most is the organizational alignment on AI adoption goals and the means to achieve them. This should be supported by employee engagement and a systematic approach to managing the portfolio of initiatives.

Summary

Establishing an AI-ready organisation depends on fundamental principles: ensuring teams possess the knowledge to leverage AI, empowering them to explore its applications, prioritising improvements in core processes, engaging in structured experimentation, and establishing effective partnerships to accelerate progress. Technology alone does not drive impact. Businesses that continuously explore how to leverage technology, empower their workforce, and approach AI adoption with discipline have the highest chances to transform technological potential into measurable, sustainable value.