GURGAON, India — May 25, 2026 — Shwetank was not looking for a new career. He had one. Nearly two decades in data analytics and data science, working with companies on business intelligence, customer segmentation, and predictive modeling. IIT Delhi in 2006. ISB Hyderabad in 2012. A comfortable corporate trajectory.

Shwetank noticed that the coaching options available in India were stuck in the past. Large batches of 15-20 students. Generic worksheets. Materials designed for the old paper SAT, even though College Board had switched to a shorter, adaptive digital format. Tutors who had never taken the test they were teaching.

So he took it himself. The Digital SAT. At 40 years old, sitting in a test center alongside teenagers.

He scored 1550. Math: 800.

"That was the moment I realized I could do this better," Shwetank said. "Not because I'm smarter than other tutors. But because I'd actually been through the current test. I knew what the adaptive format felt like. I knew where the traps were. Most coaching in India was still teaching the 2019 version of the SAT."

HE DIDN'T START WITH A BUSINESS PLAN

Shwetank began tutoring a few students in Gurgaon under the name "SAT Prep Gurgaon." No website. No marketing. Just word-of-mouth among parents in his social circle.

The first thing he did differently was diagnostic-driven instruction. Instead of starting with a curriculum ("Week 1: Algebra"), he started with a test. He analyzed each student's practice test results, identified the specific question types where they were losing points, and built a plan around those areas.

"One student was strong in math but kept losing points on word-in-context questions in the reading section," Shwetank said. "Another had timing issues specifically in Module 2 of the adaptive test. Same SAT, completely different problems. You can't fix both with the same curriculum."

His data analytics background showed. He tracked time-on-question, accuracy by topic, module-level performance. Patterns emerged that traditional coaching missed entirely.

Word spread. Students from Dubai, Abu Dhabi, Kuwait, and the US started reaching out. The name changed to SATPrepIn. The approach did not.

WHAT MAKES IT DIFFERENT

SATPrepIn does one thing: Digital SAT coaching for Indian students. No GRE. No GMAT. No IELTS. Just SAT.

The coaching is one-on-one. No batches. Every student gets a diagnostic first, and the plan is built around their specific weaknesses. If a student is strong in math, the coaching focuses on reading and writing. If the issue is timing strategy for the adaptive format, that's where the time goes.

This is harder to scale. It costs more to deliver. But the results are better because every session targets something the student actually needs to work on.

"At 1400, the difference between a 1400 and a 1500 isn't intelligence," Shwetank said. "It's execution. Systematic, repeatable processes for approaching each question type. That requires individual attention."

THE DATA ANALYST BECOMES A TUTOR

Shwetank's path from corporate data science to SAT coaching was not linear. He spent years working with data before he ever thought about education. Customer behavior analysis. Predictive models. A/B testing.

"I think like a data analyst, not a tutor," he said. "When a student gets a question wrong, I want to know why. Was it a knowledge gap? A timing issue? A trap answer? Did they rush or did they deliberate? The answer to that question determines what we work on next."

That analytical approach led to MyCollegeBook (mycollegebook.org), a practice platform Shwetank built for his students. It has 3,000 original SAT questions, 37 math skills, 13 reading and writing question types, and 6 full mock tests that mirror the Digital SAT's adaptive format.

The platform tracks what most practice tools don't: how long a student spends on each question, where they hesitate, where they rush, and whether their wrong answers correlate with longer or shorter time-on-question.

The data from MyCollegeBook has produced findings that Shwetank says no other SAT platform has published. In a study of hundreds of completed practice tests, his team found that SAT Module 2 accuracy drops 25 to 30 percent across every topic compared to Module 1. But the most interesting finding was the timing data: in Module 1, students spend 30-40 extra seconds on questions they get wrong. In Module 2, that extra time disappears. Students answer wrong questions at the same speed as correct ones.

"They're confident and wrong," Shwetank said. "Module 2 isn't harder because the math is harder. It's harder because the wrong answers are engineered to look right. Students pick them in under 20 seconds and never question it."

The study is published on MyCollegeBook's website, free for anyone to read.

WHERE IT IS NOW

SATPrepIn coaches hundreds of students across India, the Middle East, and the United States every year. MyCollegeBook is available for a one-time payment of $9.90. No subscription. No premium tier. Competing platforms charge $50 to $200 per month.

Shwetank still coaches personally. He takes on a limited number of one-on-one students and trains a small team of tutors who use the same diagnostic-driven method.

"I didn't plan to leave corporate," he said. "But when I saw how much the coaching industry in India was getting wrong, and how much the data could fix, I couldn't unsee it."

CONTACT:
SATPrepIn
[email protected]
https://satprepin.com
https://mycollegebook.org