UPSC Mains feedback
has been broken for decades.
15,000 aspirants write Mains every year. Most get feedback that says “add more dimensions.” That's not feedback — it's a stamp. We built the alternative.
The coaching test series problem
The standard way to get UPSC Mains feedback is a coaching test series. Pay ₹10,000–30,000. Submit your handwritten answers. Wait two to three weeks. Get back a stamped sheet.
The stamp might say “analytical depth lacking.” It rarely tells you which answer, which paragraph, or what a better version looks like. One evaluator handles 200+ students. The economics make specific feedback impossible.
2500 aspirants clear Mains to interview each year with roughly a million in the competition. The gap between attempt and selection almost always comes down to answer writing — not knowledge. Most aspirants know enough. They don't know how to write.
LeanPrep exists because that gap is solvable. Specific feedback. For every answer. In under five minutes.
Dhawal Raturi
Founder & CEOI'm an engineer who spent months working with UPSC rankers and aspirants — trying to understand examiner psychology and the student experience, including how coaching traps aspirants in a perpetual cycle of costly Mains Test series. I came to UPSC prep as an outsider — which turned out to be useful.
I dug deep into how UPSC Mains evaluation actually works: what examiners look for, where coaching feedback fails, which dimensions differentiate a 6-mark answer from a 9-mark answer. Then I built for that.
LeanPrep is a focused operation built on the conviction that specific feedback compounds faster than more content.
Things We Don't Compromise On
These aren't values statements. They're constraints that shaped every feature.
Every answer gets its own score and its own explanation. “Your GS3 sheet was weak” is not feedback. “Answer 3 missed the international dimension and the constitutional angle” is.
Point, Explain, Contextualize, Apply. This structure comes from analysing how UPSC examiners actually award marks across papers — not from a coaching manual. Every evaluation maps your answer against it.
GS1 is not GS4. An Ethics answer is graded on different dimensions than a Polity answer. We built separate scoring logic for GS1, GS2, GS3, GS4, and Essay rather than applying one rubric to everything.
Feedback you receive three weeks after writing doesn't change what you write tomorrow. We treat turnaround time as a product requirement, not a nice-to-have.
Every question gets a model answer built to score in the top 10% — not a generic template. Introduction with a contextual hook, body structured around the question's actual directive and themes, conclusion drawn from what was argued. The same construction logic a ranker applies on paper.
Knowing what you got wrong isn't enough. Each evaluation surfaces the enrichment that separates a ranker's answer from an average one: topic-specific keywords, data points, constitutional provisions, Supreme Court judgements, government schemes, and quotes — crisp, bulleted, ready to use.
UPSC examiners revisit themes. After every evaluation, we search across the UPSC PYQ bank and surface the past questions that tested the same topic — so you can see how the examiner has framed the same idea before and practice the pattern.
Our model answers are grounded in verified information. We don't generate facts, invent examples, or hallucinate case studies. If the AI isn't confident, it says so.
Start with 5 free evaluations
Upload any handwritten UPSC answer sheet. Get question-by-question feedback in under 5 minutes. no payment needed.
Evaluate My Answer Free →