Throughline: Teachers focus on today. The AI holds the long view.
A planning partner for primary teachers that sequences the curriculum, adapts it from a minute of feedback after each lesson, and explains every change it makes.
Better planning. Better teaching. More room for what matters.
Teachers make hundreds of decisions throughout the school day. Long before entering a classroom, they must interpret curriculum requirements, prepare resources, consider individual learning needs and anticipate how each lesson contributes to long-term educational outcomes.
For a multi-school education organisation, these challenges were amplified by differences in planning methods, fragmented resources and the growing responsibility to deliver consistently high-quality, inclusive teaching.
The opportunity was to explore whether artificial intelligence could reduce the administrative burden of lesson planning without compromising the creativity, flexibility and professional judgement that make effective teaching possible.
Throughline was developed as a design concept for an AI-supported teaching and curriculum-planning experience.
Rather than replacing the decisions teachers make, the product was designed to make those decisions easier to understand, adapt and act upon.
The challenge
Lesson planning is rarely a linear process.
Teachers work across curriculum frameworks, medium-term plans, knowledge progression documents, assessment expectations and classroom resources. These materials are often stored in different formats and adapted independently by individual schools or departments.
The challenge becomes more complicated when considering Special Educational Needs and Disabilities (SEND), varying levels of understanding and unexpected changes during lessons.
Teachers need consistency in what students learn, but flexibility in how they teach it.
This created a fundamental design tension.
How might we reduce the complexity of lesson planning while protecting teacher autonomy and improving the consistency of learning?
The solution needed to address more than document creation. It had to support the wider relationship between curriculum intent, teaching decisions and classroom reality.
Discovery and research
Before defining a solution, the project began with understanding how educators plan, deliver and adapt learning.
We explored the existing planning experience through collaborative discovery workshops, stakeholder discussions, journey mapping and reviews of educational resources.
Conversations brought together perspectives from teachers, curriculum leaders, digital learning specialists and staff responsible for supporting learners with additional needs.
This research allowed us to challenge the original assumption that the problem could be solved by simply generating lesson plans more quickly.
Instead, we began to uncover a more interconnected system of decisions, documents and responsibilities.
Research activities
01. Stakeholder discovery
Understanding organisational objectives, digital readiness and everyday planning challenges.
02. Journey mapping
Exploring how teaching preparation connects to the experiences of teachers and learners.
03. Workflow analysis
Identifying the documents, decisions and dependencies involved in planning.
04. Co-design workshops
Translating research insights into opportunities and product requirements.
What we discovered
Fragmented resources
Curriculum information, teaching materials and supporting documentation were distributed across multiple files and systems. Bringing them together required significant manual effort.
Repetitive planning work
Teachers frequently adapted existing resources, repeated planning activities and recreated materials for different classroom contexts.
Personalisation requires time
Supporting varying abilities, SEND requirements and individual learning needs introduced additional planning complexity.
Learning rarely follows a fixed plan
Unexpected disruptions, gaps in understanding and different rates of progression regularly required teachers to adjust the intended learning sequence.
Trust and control were essential
Educators needed to understand why AI recommended particular content and retain ownership over decisions affecting their students.
key Insight
Teachers didn't necessarily need AI to do more. They needed it to take responsibility for the right things.
Key insights
The research revealed an important distinction between creating teaching materials and managing the complexity surrounding them. Generating more content could introduce an entirely new burden if teachers then had to spend time reviewing, correcting and understanding large quantities of AI-produced material.
The product therefore needed to support informed decisions and curriculum continuity, not simply maximise the volume of automatically generated lessons.
Reframing the opportunity
Early exploration considered AI-assisted lesson planning as a way to accelerate the creation of educational materials.
However, the research showed that the greater opportunity sat above the individual lesson. Teachers needed to understand how different learning objectives connect, what prior knowledge students require and how changing one lesson might affect those that follow.
This prompted a shift in product direction.
INITIAL THINKING
AI generates lesson plans
A system focused on producing teaching documentation and reducing time spent writing content.
REFINED DIRECTION
AI supports learning continuity
A system that helps teachers understand, structure and adapt sequences of learning over time.
Clarity over complexity. Present the information teachers need when they need it, without overwhelming them with an entire curriculum's worth of content.
Assistance without substitution. Let AI propose, organise, and explain, while teachers can review, modify, and direct the output.
Continuity without rigidity. Maintain a clear view of the intended learning journey while allowing teachers to respond to what happens in the classroom.
Designing the experience
The product experience was structured around how educators naturally approach planning, beginning with wider learning objectives before progressively focusing on individual lessons.
Rather than asking users to navigate a complex set of disconnected tools, the interface guides them through a clear planning sequence.
01
Define the context
Select the subject, year group and planning period.
02
Establish learning goals
Set outcomes using curriculum requirements and teacher input.
03
Review the sequence
Explore the proposed progression and its reasoning.
04
Adapt to the classroom
Consider resources, constraints and learner needs.
05
Refine and use
Review the outline and prepare an editable teaching plan.
The goal was to make the planning journey feel structured without being restrictive. Each stage introduced a manageable decision, with the option to provide further context or refine AI-supported suggestions. This progressive approach reduced the need to confront every possible teaching consideration at once.
UX/UI design
The visual direction for Throughline was built around clarity, confidence and restraint.
Rather than presenting AI as a conversational interface that teachers must continually prompt, the experience uses familiar patterns: structured forms, lesson cards, progress indicators, contextual explanations and straightforward editing controls.
The identity combines deep navy, confident blue accents and generous white space to create a focused, professional environment.
The interface was designed to communicate what the system is suggesting, why a recommendation has been made and where teachers can intervene.
Small explanatory elements provide additional context without forcing users to leave the planning journey.
Visible lesson sequencing helps educators understand their current position and the broader learning objectives they are working towards.
Crucially, AI suggestions are presented as support for professional decision-making, rather than unquestionable instructions.
The solution
Throughline brings together curriculum context, learning objectives and teacher input within a structured planning experience.
Its proposed interface enables educators to move from broad educational goals towards the individual lessons they need to prepare, with AI helping to organise relevant information and explain the reasoning behind recommendations.
The product concept combines four connected capabilities.
Curriculum-aware planning
Builds proposed learning outlines around defined subject knowledge, objectives and progression requirements.
Connected learning sequences
Gives teachers visibility of how individual lessons relate to prior knowledge and future learning.
Responsive planning
Explores how teacher feedback and changing classroom requirements could inform future lesson adjustments.
Explainable recommendations
Allows educators to interrogate the reasoning behind suggested topics, content and lesson ordering.
The proposed result is an AI experience that supports the full context of planning without attempting to control the delivery of a lesson.
A product that holds the broader learning journey in view, leaving educators free to concentrate on the students in front of them.
Outcomes and impact
The project translated a broad ambition to explore artificial intelligence in education into a more clearly defined product experience grounded in the realities of teaching.
Through collaborative research and design, the team established a clearer understanding of planning workflows, identified significant user needs and developed a concept that could be discussed and challenged by educators.
The work demonstrated the value of using UX research to move beyond the possibilities of a technology and towards the needs of the people expected to use it.
3
Wireframe feedback sessions documented
2
User journeys considered
1
User journeys considered
A clearer product proposition. The opportunity evolved from generic AI content generation into an experience focused on curriculum continuity, explainability and planning flexibility.
Research-informed product requirements. Direct engagement with educators helped identify which features could provide meaningful support and where automation might introduce unnecessary friction.
A foundation for further validation. The wireframed experience provided a tangible basis for evaluating usability, refining the product and determining what should be prioritised in a proof of concept.
The wider proof-of-concept brief defined usability testing, teacher feedback, workload measurement and potential scaling as areas for evaluation. The available records do not establish verified reductions in planning time, improved pupil attainment or full product deployment. Those outcomes are therefore not claimed here.
Reflections
The best use of AI isn't always doing more.
One of the most valuable lessons from Throughline was recognising the difference between what artificial intelligence can generate and what users actually need.
In education, more output doesn't necessarily mean better outcomes.
An AI system that can create extensive lesson plans may seem efficient, but it can also add demands on teachers, who must review, understand, and adapt its recommendations.
Designing around that tension meant continually asking where automation provided genuine value and where the teacher's involvement remained essential.
The project reinforced the importance of treating AI as part of a wider human experience.
Rather than designing around a technical capability, we focused on how decisions were made, what information was needed and how educators could maintain confidence in the process.
It also highlighted the importance of designing flexibility into systems that operate in unpredictable environments.
A successful learning plan should provide direction without becoming a constraint.
