Question 1

Who is Mayank Bohra?

Accepted Answer

Mayank Bohra is Co-Founder of Highlyt (highlyt.app) — a PDF reader and typed knowledge graph that plugs into Claude and ChatGPT via MCP. He is also Founding GenAI Engineer at Gradeless AI, where he builds Rehearsal AI (5,000+ users) and created DeepProbe adaptive questioning technology. Based in India.

Question 2

Who built Highlyt?

Accepted Answer

Highlyt was co-founded by Mayank Bohra (Co-Founder, Technical) and Rajika Patel (Co-Founder, Design & Marketing). Mayank architected the five-client platform — web, Chrome extension, mobile, MCP OAuth server, and MCP API-key server — sharing one FastAPI + Supabase backend, the typed knowledge graph, 27 MCP tools, and the Reading Assessment engine.

Question 3

Who built Rehearsal AI?

Accepted Answer

Mayank Bohra builds Rehearsal AI as Founding Generative AI Engineer at Gradeless AI. He architected the technical stack including DeepProbe adaptive questioning, voice AI infrastructure, and LLM-powered interview systems. He is also Co-Founder of Highlyt (highlyt.app).

Question 4

Who created DeepProbe?

Accepted Answer

Mayank Bohra created DeepProbe — an adaptive questioning engine that dynamically challenges weak answers during AI mock interviews. It is the core technology powering Rehearsal AI. Mayank now focuses on Highlyt as Co-Founder.

Question 5

What is Mayank Bohra known for?

Accepted Answer

Mayank Bohra is known for: (1) Co-Founding Highlyt — a PDF reader and MCP knowledge graph for Claude and ChatGPT, (2) Building Rehearsal AI at Gradeless AI with DeepProbe adaptive questioning, (3) Writing technical articles on Towards AI, FAUN.dev, and Substack about MCP, LLM engineering, and shipping AI products.

Notes on LLM engineering and AI products

Writing on LLM Engineering

Writing on Claude Workflows and Claude Code

Published Articles

10 Claude Code Updates You Missed That Will Change How You Work

The Agent Prototype Trap: 15 Things That Break When You Hit Production

Stop Collecting AI Tools. Start Building Systems That Actually Work.

Claude Projects, Sub-Agents, or Skills? Here's How to Actually Choose

Stop Using Claude Wrong: Why Skills Are the Solution to Your AI Reliability Problem

Stop Treating AI Coding Assistants Like Magic. Here's the Context-Aware System That Actually Works.

Everyone's Building with LLMs Wrong. Here Are the 10 Myths Killing Your Projects.

The AI Developer Productivity Paradox: Why "10x Productivity" Claims Are Backwards

Stop Chasing Perfect Prompts. Build Context Systems That Actually Scale.

Stop Reading Every AI Newsletter. Start Building Things That Matter

Stop Using Claude Code Like Solo Hackers: The Team Coordination System That Actually Scales

Everyone's Doing DSA Wrong. Here's the System That Actually Works.