Phase 1
Identifying Opportunities
In the first phase, we work with your team to identify where Frag-Maria can create real value. We look at use cases, channels, languages, and data sources, and prioritize initial ideas based on impact and effort. This creates a clear overview of where a sensible starting point lies and which KPI hypotheses can be made measurable later on.
Phase 2
Refine your idea
In this phase, a good idea is refined into a precise MVP. The use case is condensed to its essentials, initial edge cases are considered, and we examine whether an existing solution or partner integration is the right approach. Simultaneously, we clarify key questions regarding data, roles, logging, opt-in, security, and compliance, ensuring that the launch is not only swift but also sustainable.
Phase 3
Define Priorities and Roadmap
Now we decide what Maria will be trained and focused on first. Together with your subject matter experts, we prioritize the most important jobs-to-be-done: What needs to be solved immediately, what follows in the next iteration, and what will be deliberately put on hold for now? This results in a roadmap with clear targets, skills, and priorities for the initial expansion stages
Phase 4
Build, test, improve
In this phase, the MVP is technically implemented. We build flows, prompts, knowledge bases, and relevant integrations, and test Maria with real cases in staging. The goal is a functional, secure, and practical use case with clear fallbacks, clean handover, and a robust go-live basis.
Phase 5
Go live and fine-tune
Maria is now going live on the defined channels. We monitor the initial deployments, measure key metrics such as TTR, FCR, handover rate, satisfaction, and conversion, and specifically fine-tune tonality, routing, and response quality. This transforms the MVP into a real live operation with initial quick wins and a clear foundation for Sprint 2.
Phase 6
Empowering employees
Finally, we empower your team to confidently and effectively continue using Maria in their day-to-day work. This includes training, documentation on flows, data sources, roles and KPI logic, as well as clear guidelines for maintenance, escalation and further development. At the same time, we retain technical responsibility for 1st level support – including maintenance, patches, and security and compatibility updates.
