One Company’s Blueprint for Taking a Whole Engineering Org Agentic

Key Takeaways

• Ida Infront targets 2x development speed by year‑end using Kilo
• Structured three‑workshop rollout trains AI ambassadors across 70 seats
• Early pilots report jaw‑dropping productivity gains from agentic coding
• AI agents shift developer roles, enabling domain‑specific automation
• Customer ARN sees faster case processing via AI classification built by Ida

Pulse Analysis

Legacy government software often stalls innovation because compliance, risk, and decades‑old codebases limit experimentation. Yet the pressure to deliver faster, cheaper solutions is mounting, especially as AI tools become mainstream. Ida Infront’s decision to standardize on an agentic engineering platform reflects a broader trend: enterprises are moving from ad‑hoc AI assistants to fully autonomous coding agents that can orchestrate tasks, manage dependencies, and even propose architectural changes. By committing to a single tool and embedding it in a structured onboarding program, the company sidesteps the chaos of fragmented tool adoption and creates a repeatable pathway for scaling AI benefits.

The three‑phase workshop model—covering fundamentals, iterative debugging, and non‑technical alignment—acts as a catalyst for cultural shift. AI ambassadors, selected from each team, prepare repositories with agent‑ready metadata, ensuring that the Kilo agents have the context they need to operate effectively. This hands‑on approach not only accelerates skill acquisition but also surfaces developer concerns early, turning resistance into constructive feedback. Early indicators show developers moving from “assistant‑only” speed gains to a redefined role where agents handle routine or domain‑specific tasks, freeing engineers to focus on higher‑value problem solving.

Beyond internal productivity, Ida Infront is leveraging the same agentic technology for client outcomes. The AI‑driven classification feature deployed for Sweden’s ARN board cut manual processing time dramatically, illustrating how agentic code can translate into tangible service improvements. As other subsidiaries eye the model, the ripple effect could reshape procurement standards for public‑sector IT, prompting vendors to embed autonomous agents into their offerings. For the broader market, Ida Infront’s experience signals that even the most legacy‑laden organizations can achieve double‑digit speedups when AI adoption is disciplined, organization‑wide, and tied directly to product innovation.