Sri Lankan technology firm 99x hosted ELEVATE 2026, bringing together over 150 employees to develop practical applications of AI agents in engineering workflows, signalling an industry shift beyond simple code generation.
99x, a global technology company based in Sri Lanka, recently organised ELEVATE 2026, an internal hackathon designed to explore how artificial intelligence agents can be integrated into real-world software engineering practices. The event drew participation from more than 150 employees organised into 30 teams, each tasked with developing AI agent solutions addressing challenges specific to their project domains.
The hackathon reflected a broader evolution in how the technology industry approaches AI in software development. According to Yohan Gunathilaka, Senior Manager of Software Engineering at 99x, the focus has moved beyond using AI for basic code generation. Instead, teams were encouraged to consider AI agents as integrated components within engineering workflows, examining when to trigger them, what outputs they should produce, and critically, where human decision-making remains essential. This represents a more sophisticated approach to agentic development than simply using AI to work faster.
The company's Xianix platform serves as the conceptual foundation for this approach, combining specialised AI agents, Forward Deployed Engineers (FDEs), and engineering guardrails within delivery pods. Core agent capabilities include requirement analysis, technical design, pull request review, test planning and release management. Importantly, human engineers retain accountability for decisions at each stage of the process.
99x has emphasised that its AI agents are intended to augment rather than replace human engineers. As these systems take on routine analysis, preparation and initial review tasks, engineers must develop new competencies in evaluating AI output, managing associated risks and making informed decisions. According to Asitha Weerasinghe, Senior Technical Architect on the ELEVATE organising team, the conversation within the company has shifted significantly from simply making teams comfortable with AI assistance to exploring how agents can function across the entire project lifecycle while maintaining engineering discipline and customer accountability.