November 16, 2026
Read time: 5 min.

Sombra Has Changed Its Approach to Development: How the Company Is Advancing AI-First and AI SDLC 

Sombra already applies an AI-first approach in the daily work of its engineering teams. The company is developing AI SDLC — an approach to the software development life cycle in which AI performs not only individual tasks but entire functions in the product creation process. This changes how work is organized, how teams are composed, and how responsibility is divided between engineers and AI agents. The international IT company brings together more than 400 specialists, has a presence in Ukraine, Colombia, and India, and works with clients in North and Latin America, Europe, and the Middle East.

The next goal is to scale the approach that already works. By the end of 2026, Sombra plans to cover all of the company’s engineers with AI-first practices and, for teams under its management (managed teams), to reach the target level of AI SDLC Level 3. This is not about starting to use AI, but about spreading practices and maturing them across the entire development cycle.

AI SDLC: A New Distribution of Functions in the Team 

Sombra’s focus is not on separate AI assistants for writing or reviewing code, but on organizing development according to AI SDLC. AI can perform entire functions and take on tasks that were previously assigned to separate roles: from working with requirements and design to implementation, testing, and verification of results. For engineers, this means working by new rules: deciding which functions to delegate to AI, how to combine them into a shared process, and how to verify the result. 

Team composition is also changing: what matters is not only the set of specializations but also the distribution of functions between people and AI agents. At the same time, responsibility for context, quality, and final decisions remains with people. 

As a next step, Sombra is considering the AI Software Factory model. In it, a human defines the requirements, scope of work, and key constraints, while AI agents can carry out a significant part of design, development, testing, and verification. Control of the result and the final decision on releasing the software remain with a human. 

Therefore, the development of AI-first at Sombra covers not only tools but also new formats of teamwork. Engineers need a broader understanding of the product and the ability to organize task execution beyond a single narrow technology — taking into account business context and client needs. 

How Sombra Prepares Engineers for AI-First 

Sombra combines training, practical engineering scenarios, and assessment of AI-first skills. At the end of the third quarter, the company had 17 developers with Claude certification; the stated goal for the fourth quarter is 25. 

The company’s specialists also hold Anthropic, Google, Microsoft, and AWS certifications. The key emphasis of the training is the ability to set a task for AI, integrate it into the development process, and evaluate the quality of the result. 

AI SDLC Hackathon: Testing the Approach in Practice 

Sombra held an internal AI SDLC Hackathon as a practical test of the AI-first approach to development. The goal was to examine how engineers use AI to complete a development task within defined requirements and constraints.

Results of the first hackathon 

  • 19 participants completed and submitted their projects
  • 17 of 19 met all functional requirements
  • 16 of 19 stayed within the budget of $10

For Sombra, the hackathon was not only an internal competition but also a practical experiment in AI-assisted development. The company explored at which stages AI can speed up project delivery, where active engineer involvement is needed, and how to control the quality and cost of such development. 

Sombra plans to continue AI SDLC hackathons as a way to test new development formats and build AI-first competencies. 

Why Sombra Is Developing AI-First 

Sombra considers AI-first important not only as a way to automate individual operations. This approach makes it possible to rethink how a team creates a product: which functions are performed by people, which by AI, and how to organize their interaction around business needs. 

The company expects that the development of AI SDLC will help shorten the time from requirement to finished solution, form teams more flexibly for tasks, and test product ideas faster. These are expected results of scaling the approach, not universal indicators already achieved on all projects. 

At the same time, speed does not replace engineering responsibility. Critical thinking, product understanding, and quality verification remain the foundation of the work: it is important to assess not only whether a solution works, but also whether it fits the specific business context. 

What’s Next 

AI is already part of the daily work of Sombra’s teams. The next stage is to adopt new formats of development and of human–AI-agent interaction, scale proven practices, and develop AI SDLC. 

A separate focus is to proactively offer clients new approaches to delivering their tasks and to adapt collaboration formats to new needs. For Sombra, this is a continuation of work with an approach that is already in use, and an opportunity to expand its application in client projects.

Share:
Sombra Has Changed Its Approach to Development: How the Company Is Advancing AI-First and AI SDLC 

Sombra already applies an AI-first approach in the daily work of its engineering teams. The company is developing AI SDLC — an approach to the software development life cycle in which AI performs not only individual tasks but entire functions in the product creation process. This changes how work is organized, how teams are composed, […]

https://itcluster.lviv.ua/wp-content/uploads/2026/11/2026.08__315.jpg
Copied!