Capgemini Exceller Technical Assessment 2026: AI + Technical

The Capgemini Exceller technical assessment is now two multiple-choice sections, not one: an AI Literacy section and a classic technical section. If your prep only covers pseudocode and DSA, you are ready for half of it.

This is the biggest change from older Capgemini test guides, and it is also an elimination gate. Clear it and you reach the coding and interview stages. Miss it and the process ends here, whatever your scores elsewhere. The full journey is mapped in the Capgemini Exceller framework guide; this page zooms into the technical module alone.

What the Exceller technical module tests

Capgemini describes Exceller on its technical graduates page as its route for hiring emerging engineering talent. The technical module is where it checks baseline engineering literacy before any coding round. The shape shared with candidates:

Section	Focus	Format
AI Literacy	Generative AI, prompts, responsible AI	Multiple-choice
Technical	DSA plus CS fundamentals	Multiple-choice

Two sections, roughly twenty questions each, inside a combined time window. Treat the two-section MCQ structure as reliable and confirm exact counts with your placement cell, since Capgemini notes in its recruitment process guidance that steps vary by drive.

Section 1: AI Literacy and generative AI

AI Literacy is new, and that makes it low-competition. Few candidates have prepped for it, so a small amount of focused study goes a long way. The topics reported in the framework shared with candidates:

• Generative-AI foundations: what foundation models and large language models are, at a conceptual level.
• AI capabilities and limitations: what these systems do well, and where they fail.
• Prompt basics: prompt structure, context setting, and what makes a prompt clear.
• Advanced ideas at an awareness level: retrieval-augmented generation, agentic AI, and multi-step AI workflows.
• Responsible AI: output validation, bias awareness, and safe-use practices.

You are not asked to build a model. You are asked to reason about these concepts. A candidate who can explain why a language model might produce a confident but wrong answer is answering the kind of question this section favours.

Section 2: technical (DSA and CS fundamentals)

The technical section stays close to what Capgemini has always tested, so older prep still applies here. The high-frequency areas:

• Programming logic: variables, data types, input and output logic, and pseudocode interpretation.
• Data structures and algorithms: arrays, strings, hashing, searching, sorting, and time and space complexity.
• Software-engineering fundamentals: OOP, DBMS basics, SQL queries, REST APIs, HTTP methods, and version-control or Git concepts.
• Modern engineering awareness: client-server architecture, cloud basics, networking fundamentals, and cybersecurity basics.

Pseudocode tracing sits at the centre of this section. You read a block of logic and answer what it outputs or where it breaks. Our Capgemini pseudocode MCQ practice is a direct fit for building that skill.

Sample question types (illustrative, not actual papers)

The examples below are illustrative of the question types, written by FACE Prep to show the format. They are not reproduced from any Capgemini paper.

• AI Literacy type: given a short scenario, pick which limitation of a language model it demonstrates. Checks whether you understand model behaviour, not memorised trivia.
• Prompt type: choose the better-structured prompt for a stated task. Checks context-setting and clarity.
• DSA type: given a loop or recursion snippet, select the output. Checks logic tracing under time pressure.
• Fundamentals type: pick the correct SQL clause or the right OOP concept for a described behaviour. Checks recall of core definitions.

How the two sections differ, and where to spend time

The two sections reward different preparation, so split your effort by return. The technical section is the more predictable of the two. Its patterns repeat across drives, so earlier practice compounds. If you prepared DSA basics and CS fundamentals for any other company test, most of that work carries over here. The marginal hour is best spent on your weakest topic, whether that is complexity analysis, SQL, or OOP behaviour.

AI Literacy is different. It is new, so there is less shared prep floating around, and a small focused study block moves your score more than the same hour on already-strong DSA. Read a plain-language explainer on how language models generate text, one on how retrieval adds context, and one on responsible-AI basics. That coverage handles most awareness-level questions.

Timing is the quiet decider. At roughly one minute per question, there is no room to stall on a single hard item. Answer what you know first, flag the uncertain ones, and return with the time you saved. Since candidate reports describe no negative marking on the section, a blank only costs a possible mark. Attempt everything before the timer ends.

One point on mindset. The technical module is an elimination gate, so the goal is to clear the bar reliably, not to chase a perfect score in one section while the clock runs out in the other. Balanced accuracy across both beats brilliance in one.

How to prepare for the technical module

Prioritise by frequency and by elimination risk. Start with DSA basics and CS fundamentals, since they carry the most questions and the most predictable patterns. Layer AI Literacy on top: a few hours on generative-AI concepts, prompt structure, and responsible-AI ideas covers most of what this section asks. Drill pseudocode tracing to build speed, because a one-minute-per-question pace leaves no room for slow reading.

For the wider context of how Capgemini structured its aptitude and reasoning rounds in earlier years, the discussion of Capgemini aptitude questions is a useful companion read. Map your weakest of the two sections first, then come back to the framework guide to sequence the debugging and AI-assisted coding stages that follow this one.