Ishaan Gangwani

Ishaan Gangwani did not start as a founder, he started as an annoyed teen researcher fighting LaTeX errors and clunky tooling while co authoring papers across time zones. A professor at the National University of Singapore introduced him to fellow student researcher Aayam Bansal, and within weeks they were co writing work that would eventually get rejected from a major conference, the Association for Computational Linguistics.

That rejection flipped the frame from academia to product. Instead of taking the signal as a verdict on their ability, they treated it as user research on the broken workflow around scientific writing and review. The insight was simple but sharp: every researcher fights the same bottlenecks in literature review, experiment design, and drafting, yet the software layer is still tied to legacy tools like LaTeX and manual copy paste, which is where a focused AI system could be a 10x improvement.

Their first shot at fixing the problem was InkVell, an AI native operating system for scientific research that they built from Calcutta and Pune while still teenagers. InkVell turned their frustration into a structured environment for drafting and managing papers, and it attracted enough traction and investor belief that they raised roughly $1.4 million and won a slot in Y Combinator Winter 2026, making them some of the youngest Indian founders to ever enter the program.

Post YC, the pair doubled down on the same user, but a sharper wedge, by launching Synthetic Sciences from San Francisco as an AI co scientist for end to end research. Synthetic Sciences focuses on concrete workflows like reviewing studies, planning experiments, and running analyses, and its Biology mode has already posted a 92 percent score on the BixBench benchmark, which YC itself highlighted as state of the art. That benchmark performance, combined with the YC W26 brand and a $500,000 YC check inside a roughly $1.5 million total raise, pulled in serious researchers as early adopters and shifted outside perception from novelty teen founders to credible infrastructure for labs.

1. Start with a painful personal bottleneck, then generalize only after you have felt the friction multiple times in real work. Ishaan and Aayam did not brainstorm AI startup ideas in a vacuum, they tried to ship research, hit the wall on LaTeX and review cycles, and then built exactly the tools they wished they had, first with InkVell as an operating system for papers, then with Synthetic Sciences as a co scientist that lives inside the research loop.

2. Use legitimacy ladders to neutralize the age discount with investors and users. Instead of leaning on being young, they stacked credible signals like research collaborations, NUS introductions, and serious benchmarks, then converted those into institutional validation through Y Combinator W26 and a BixBench verified score. Each rung, from the early InkVell raise to the YC check and public benchmark posts, made the next investor conversation about rigor and results instead of whether two teenagers could be trusted with real science.

3. Anchor distribution around where working scientists already are, and make every proof of work a piece of content that travels. Synthetic Sciences ships directly into familiar researcher workflows by helping with literature review, study design, and experiment execution, then uses public metrics like the 92 percent Biology score and YC feature posts as shareable artifacts for LinkedIn, startup press, and word of mouth among labs. The playbook is simple: solve a deep workflow, publish concrete numbers, and let skeptical experts spread the story because the tool actually reduces their daily pain.