Registration closes 25 Sep, 09:00. Two registrations, both mandatory. Discipline: Computer Science, plain — it sits in the unsorted block after "Welding". Nothing else matters until this is done
Interview, which may include an extempore · pre-employment medical
The structure is a planning model, not a published fact: L&T publishes nothing beyond the two-line description. It is [LIKELY] — matched by two independent accounts (2020 and 2026) — but cohorts have seen 100 to 130 questions. See research/03-pattern-synthesis.md §0 and §2 for the competing accounts and why this one was chosen.
The library
Start here
File
What
01-REGISTRATION-CHECKLIST.md
Both registrations, documents, discipline choice, slot booking, test-day setup
The conflict table. What is known, what is not, and what is merely believed
research/04-qa-report.md
What was checked, what broke, and what was fixed — including my own mistakes
research/05-source-audit.md
How each source and link was checked, and what that changed — including that the observed questions are ~18 years old
data/audit_links.py
Re-runnable tool: fetches every link and says what kind of page it is
data/
analyze.py, weightage.json, the observed-question CSVs, and four verifiers
PROGRESS.md
Build state, decisions, open questions
Three things to know before you trust any of this
1. The 130-question structure is likely, not certain. The 2026 source was a Glassdoor page I could not open; a follow-up search found an independent Nov 2020 account that gives the identical split (120 min, technical 40 + 30 verbal + 30 logical + 30 quant). Two sources six years apart is much better than one. But other cohorts saw 120, 111 or 100 questions, so what is stable is the shape — one shared clock, no negative marking, free navigation, sectional cut-offs, a branch-set technical block — not the exact counts.
2. Register under plain Computer Science, and the technical section simplifies. The live dropdown offers it although the official PDF does not list it, so there is no variant layer to read — the shared core is the syllabus. That said: there is no public first-hand account of a CS-stream PGET technical paper at L&T Limited. AmbitionBox's entire 91-question L&T Limited GET corpus is civil and electrical. The technical material is built as core-plus-variants specifically to survive that gap, and 03-TEST-STRATEGY.md §5 carries a contingency for opening the domain section and finding thermodynamics.
3. No topic is High confidence, and 15% of the paper sits on Low. L&T publishes no syllabus, and every observed quantitative, logical and technical question comes from one paper that is probably 18 years old (research/05-source-audit.md). Low-confidence topics are capped at Tier 3 so the plan never tells you to master something the evidence cannot confirm is examined.
Reproducing the analysis
python3data/analyze.py# regenerate weightage.json and analysis-tables.md
python3data/verify_cheatsheet.py# 99 numeric claims in the cheatsheet
python3data/verify_observed.py# the observed pool's answers
python3data/verify_keys.py# every question bank: key vs question vs explanation
python3data/verify_mock.py# the mock's computable answers