When the Progress Card Reads the Classroom: AI in the Preparatory Stage

• 2026-07-21 12:00:00

• Parakh NCERT

• By Indrani Bhaduri

A teacher asks three Grade 4 learners to measure the same table. Aanya writes the correct answer but cannot explain how she found it. Kabir writes the wrong answer, checks the ruler and corrects his method. Meher notices that two groups have different answers and discovers that one group did not start measuring from zero. Aanya has the correct answer. Kabir recognises and corrects a mistake. Meher identifies why the answers are different.Who has demonstrated the strongest understanding? A conventional test may reward only Aanya. The Holistic Progress Card can capture the reasoning of all three. AI can help the teacher record and organise such learning moments before they are forgotten.
 

The Preparatory Stage Holistic Progress Card is designed to capture such differences. It moves assessment beyond marks and written examinations. It asks teachers to connect classroom activities with curricular goals, competencies and learning outcomes. It includes performance rubrics, observational notes, self-assessment, peer feedback, caregiver observations and an annual summary.The ambition is significant. So is the documentation involved. A teacher may notice several meaningful learning moments during one activity. Some are written down. Others remain in notebooks, worksheets, photographs or the teacher’s memory. By the time the annual summary is prepared, the final product may be visible, but the learner’s journey has faded. Can artificial intelligence help HPC remember how a child learned rather than merely what the child produced? It can. AI can analyse evidence, review patterns over time and predict areas where a learner may need support. However, these predictions should be treated as prompts for further observation, not as fixed labels or final judgements. The teacher must interpret them in the learner’s social, emotional and classroom context.
 

The Preparatory Stage HPC is not simply a report card. Its value lies in the relationship among different kinds of evidence. Attendance may explain why a learner missed part of a concept. An interest may provide a route into a new activity. Self-assessment may reveal confidence or hesitation. Peer feedback may show how the learner participates in a group. A caregiver’s observation may reveal abilities that are less visible at school. A teacher’s note may capture the moment when a learner changes an ineffective strategy. Each piece provides only a partial view. Together, they present a fuller account of learning.
 

The HPC already provides a meaningful framework for assessing a learner’s progress. It is guided by classroom interaction, teacher observation and professional judgement. AI can provide additional support. It can transcribe brief observations and organise work samples. It can connect evidence with competencies and trace progress over time. It can also analyse patterns in a learner’s performance. Based on these patterns, it can suggest suitable activities, follow-up questions and areas for further observation. It can identify gaps in the available evidence and prepare draft feedback. The teacher must review these suggestions before using them. The purpose is to make planning, documentation and evidence management more efficient.
 

Part A of the HPC records information about the learner. It includes attendance, interests and emotional experiences. It also includes peer feedback, caregiver observations and resources available at home. The teacher can record any additional support required by the learner. These details may appear to be background information. However, they can help the teacher understand classroom performance. For example, Rohan’s participation in reading activities may decline over three weeks. AI can compare this change with his attendance record. It may show that Rohan missed several lessons based on the same text. The teacher can then revisit the activity before viewing his hesitation as a learning difficulty. AI shall identify a relationship that deserves attention and also establish a cause. Rohan’s absence may explain his hesitation, but so might unfamiliar vocabulary, discomfort with reading aloud.
 

The same principle applies to interests. The HPC invites learners to identify what they enjoy. These interests may include reading, gardening, music, art, craft, games or cooking. AI can connect this information with classroom planning. For example, Jasleen may frequently choose leaves, seeds and other natural materials. AI may suggest using these materials for classification, grouping or measurement activities. Her interest in nature can become a pathway to learning mathematics and language. However, Jasleen should not be labelled a “nature-oriented learner”. An interest profile can be created using AI for promoting multiple learning opportunities. It should not place a child in a fixed category. Children’s interests can change over time. Past preferences should not determine future learning experiences. Context should help teachers ask better questions. It should never be used to predict a learner’s limits.
 

The HPC asks children to reflect on their emotions and behaviour. Can they express their feelings? Can they remain calm during difficult situations? Can they understand how their friends feel? Can they respect different ideas and help others? AI can organise these responses across activities or terms. A learner may regularly report enjoying classroom tasks. However, the same learner may feel uncomfortable asking for help. This pattern can encourage the teacher to examine whether the classroom feels safe for help-seeking. A response such as “Sometimes” should not become a behavioural score. A change from “Yes” to “Sometimes” should also not be treated automatically as a decline. The learner may have developed a more realistic understanding of the statement. What appears to be weaker performance may actually reflect greater self-awareness.AI can compare responses. It can fully understand the experience behind them. If assessment values reflection, should a child’s uncertainty be treated as poor performance? Or should it be recognised as evidence of deeper thought?
 

Part B of the HPC contains formative assessment frameworks for Language Education, Mathematics Education, The World Around Us, Art Education, and Physical Education and Well-being. For each curricular area, the teacher identifies curricular goals, competencies and learning outcomes. The teacher then plans an activity and decides how the learning will be assessed. This is one of the most useful points at which AI can enter the process. AI can help the teacher examine whether the proposed activity actually produces evidence of the intended competency. Suppose a mathematics competency expects learners to compare different problem-solving strategies. The teacher proposes a worksheet containing ten multiplication questions. The activity may check whether learners can calculate correctly, but it may reveal little about how they select or compare strategies. An AI-supported planning tool could ask: Will learners explain their methods? Does the task allow more than one approach? What will the teacher observe? It may suggest asking learners to solve one problem in two ways and explain which method they found more useful. The teacher may accept, modify or reject the suggestion. 
 

Consider a language activity in which children listen to a story and create a different ending. The teacher wants to assess awareness, sensitivity to characters and creative expression. Before the activity, AI could help the teacher frame observation questions. Can the learner identify the main events? Can the learner explain how a character might feel? Does the new ending remain connected with the story? Can the learner explain why the ending was changed? How does the learner respond to a classmate’s interpretation? During or immediately after the activity, the teacher could record a brief voice note: “Kavya identified the main events and suggested that the character return to help the village. When asked why the character had left, she repeated the event but did not explain the character’s feelings.” The AI could transcribe the note, attach it to the activity and organise the evidence under awareness, sensitivity and creativity. It might draft the following feedback: “Kavya understands the sequence of events and offers an imaginative alternative. She may now be encouraged to explain characters’ actions and feelings using evidence from the story.” This is more useful than “Good work” because it identifies what Kavya demonstrated and what she should attempt next. 
 

In mathematics, Two children may reach the same answer through different reasoning. One may count each object. Another may form equal groups. A third may use a memorised procedure without understanding why it works. If assessment records only the final answer, these differences disappear. A teacher could photograph the children’s work or record short explanations of their methods. AI could organise these records chronologically. This evidence can guide the next lesson. One learner may require a more complex application. Another may need concrete objects or an opportunity to explain the relationship between the procedure and its meaning.AI can identify that two methods differ and recognise which difference matters educationally. 
 

In The World Around Us, a class may observe how seeds change over a week. Learners may draw the plant, measure its growth, record changes and discuss the conditions required for germination. The evidence is distributed across several days. AI can organise dated drawings, photographs and teacher notes. It may identify that a learner recorded visible changes carefully but did not distinguish between observation and prediction. It could prompt the teacher to ask: What did you actually see? What do you think will happen tomorrow? What makes the two statements different? Such prompts can strengthen formative assessment. They help the teacher gather missing evidence before deciding a performance level.
 

The HPC asks teachers to assess awareness, sensitivity and creativity. Performance is described at three levels: Beginner, Proficient and Advanced. AI can help teachers develop clear and observable descriptors for each level. In a story activity, a beginner may identify some main events with support. A proficient learner may arrange the events independently. An advanced learner may explain how one event influences another.AI can support the assessment. It can examine originality, relevance, flexibility and the learner’s ability to develop an idea. It can compare the learner’s present work with earlier work. This helps the teacher identify growth over time. It can also retrieve examples that support a performance level. Such support can make rubric-based assessment more consistent. It can reduce dependence on memory and help teachers avoid judging only through neatness, confidence or fluency. The teacher can then review the evidence and finalise the performance level. In this way, AI can make the assessment of diverse abilities more systematic, transparent and evidence-based.
 

A caregiver may also report that a child reads confidently at home. The teacher, however, may observe hesitation at school. AI can connect these observations. It may suggest that the child is more comfortable with familiar stories or a home language. It may also indicate that the child needs support while reading before a group. By bringing different perspectives together, AI can strengthen the holistic character of the HPC. It can reveal patterns that may not be visible in one source alone. It can also help the teacher identify areas that require further observation. In this way, AI makes self-assessment, peer feedback, caregiver observations and teacher evidence more connected and meaningful.
 

The HPC asks teachers to record observational notes. It also asks them to identify the challenges faced by a learner. Teachers must explain how the learner overcame these challenges. Suppose a teacher records: “Dev identified common shapes. However, he treated a square and a rectangle as completely different. He then constructed both shapes using matchsticks. This helped him notice that both shapes have four right angles. He also explained that a square has four equal sides.” AI can organise this observation into structured feedback: “Dev compares common shapes using their properties. Initially, he could not recognise the relationship between squares and rectangles. Using matchsticks helped him examine their sides and angles. He can now be encouraged to compare shapes that belong to more than one category.” This feedback presents Dev’s learning clearly. It identifies the challenge and the support provided. It also suggests the next learning step. The AI-generated feedback must be connected with classroom evidence. Each statement should be based on a dated observation or work sample.
 

Part C requires subject-wise and overall summaries based on performance throughout the academic year. This is where AI may offer the greatest administrative support. It can retrieve observations, work samples, rubric descriptors and learner reflections from different periods. It can then prepare a provisional narrative showing change over time: “During the first term, Sana required prompts to explain her mathematical strategies. In later activities, she began using objects and drawings to justify her answers. She now compares two methods but requires further opportunities to explain which method is more efficient.”The statement records a learning journey. It does more than classify Sana as proficient. But an annual summary generated from available data may reproduce the gaps in that data. Digitally stored work may receive greater attention than an important classroom conversation that was never recorded. Recent observations may overshadow earlier ones. Frequently assessed competencies may appear more important simply because they produced more data.AI can retrieve what has been documented. The annual summary can remain a professional judgement supported by AI-organised evidence. AI can bring together rubric levels, classroom observations and work samples from different contexts. It can help the teacher recognise that development is not always linear. A learner may work independently in one situation but need support in another. AI can preserve this variation in a developmental narrative instead of reducing performance to a single average level.
 

At the Preparatory Stage, AI can make the Holistic Progress Card more useful. It can organise evidence collected throughout the year. It can help teachers track progress and identify learning needs. It can also prepare clear developmental feedback. AI can connect classroom observations, learner reflections, work samples and rubric descriptors. It can reduce the burden of documentation. AI can make the HPC process more efficient for teachers. It can reduce repetitive documentation and organise evidence from different sources. It can track each learner’s progress across the academic year. It can also identify strengths, learning gaps and areas that require further support. This gives teachers more time to plan meaningful activities and provide timely feedback. AI can thus turn the HPC into a practical tool for improving learning!