How Public Institutions Scale Expertise

Through a multi-year agreement with Qatar's Ministry of Communications and Information Technology (MCIT), Scale is one of the partners contributing to AI adoption across government entities. The work spans sectors that operate differently from one another, and AI is being put to use across all of them.

The first post in this series described how that work was structured. Here we look at what it is doing in practice.

Inside Legislative Drafting

Drafting legislation is among the most consequential work any government does. Each clause must be checked against existing law, mapped within the national legal hierarchy, and benchmarked against regional and international legislation before it can move forward. The findings must be documented in standardized formats, conflicts flagged, and gaps identified, all before legislator judgment can even be applied.

AI is now built into that workflow. Draft clauses are automatically mapped against the relevant body of law, while potential conflicts and redundancies surface earlier in the process. Every suggestion moves through a human-controlled review interface where legislators edit, approve, or reject. Those decisions feed directly back into the system, improving its accuracy over time.

Legislative drafting is ultimately a process of coordination, research, and judgment. AI changes the speed at which this supporting work can be completed, allowing legislators to spend more time on the decisions that require their expertise.

The impact also extends beyond the original workflow. The capabilities built for legislative drafting, including document understanding, legal retrieval, and structured annotation also support the full spectrum of legislative work.

Inside the Classroom

The mentioned solution is being developed in collaboration with Qatar's Ministry of Education and Higher Education, through the E-Learning and Digital Solutions Department, which leads the educational and functional requirements, curriculum alignment, implementation framework, and evaluation of educational outcomes.

Teaching at scale runs into a hard constraint: lesson preparation, grading, administration, and planning all compete with instruction for the same finite hours, and personalized attention is consistently what gets deprioritized. The ratio of teachers to students makes this difficult to resolve through resourcing alone.

For students, an AI tutor aligned to the national curriculum offers a study mode for concept mastery and a practice mode that walks through problems step by step. Available outside school hours and designed to be judgment-free, it provides the kind of guided support that would otherwise require additional tutoring.

For teachers, an intelligent assistant automates lesson planning, homework creation, and routine administrative tasks, resulting in estimated time savings per lesson. A real-time dashboard surfaces insights on student gaps and engagement patterns, giving teachers the information to shift from standardized instruction toward targeted, high-impact teaching.

For administrators, a national-level panel provides visibility into performance across schools, grades, and teachers, enabling data-driven decisions on curriculum and resource allocation.

The platform is built to give students more consistent support, to free teachers to focus on higher‑impact work, and to provide system leaders with the structured data they need to plan at scale.

Throughout, the platform maintains appropriate data-protection measures, role-based access controls, human oversight, and continuous evaluation of accuracy, safety, and educational impact.

Adaptive and Personalized Experience

In legislative work, the multiplier is compression: weeks of research, comparison, and documentation collapsed into a much shorter workflow. Legal experts still decide what the law should say, but they reach that point with more context and less manual overhead. In education, the multiplier is coverage: one teacher can reach more students with more individual attention. Teachers still decide how to support a struggling student, but they have better visibility into who needs help and more tools to provide that support.

Both cases put AI inside expert workflows, where the users are legal researchers, teachers, and administrators. The next post in this series turns to citizen-facing tools: a national recruitment platform, a cultural discovery tool, and a nationwide up-skilling program, all designed to be used without technical expertise.