Manish Tula

AI Engineer & Researcher

Software should
feel empowering

My Core Beliefs

relentlessly
pursue
clarity.
Code must
be poetic.
Design for
moments
"Standard systems fail to adapt to real-time cognitive load."

Selected Work

MindTrace AI

An extensive research initiative investigating the intersection of LLMs and cognitive load. Instead of just asking if AI helps students, MindTrace AI measures how it alters their mental frameworks. By developing custom tracing algorithms, we analyzed real-time learning patterns to publish a highly-regarded preprint. This isn't just an app; it's a new lens on educational psychology and machine learning.

ResearchCognitive ScienceLLMs

Neurova

Founder & Lead Eng

As Founder and Lead Engineer, I engineered a robust behavioral intelligence engine from the ground up. Neurova bridges the gap between raw data and human behavior using advanced GenAI and Python. I architected a scalable, containerized microservices infrastructure with Docker and LangChain, processing complex user logic inspired by Stanford's CS106A methodologies. It’s an engine designed to understand, not just compute.

Tech Stack

  • ▹ Python & GenAI
  • ▹ LangChain
  • ▹ Docker Microservices
  • ▹ Behavioral Logic

MindBloom

A global, youth-led mental health sanctuary. MindBloom isn't just another telehealth app—it's a community-driven ecosystem. By focusing on highly performant, accessible web design, I scaled the platform to actively support over 200 global users. It features real-time interactions, secure data handling, and an aggressively optimized frontend that ensures a seamless experience for vulnerable users when they need it most.

Next.js
*** 08/28/2026 ***

Vyntra

Vyntra pushes the boundaries of localized, low-latency conversational AI. What started as an experimental Python script evolved into a highly optimized voice assistant serving 98 active users daily. By stripping away bloated dependencies and writing hyper-efficient audio processing logic, Vyntra delivers near-instantaneous responses, proving that powerful AI doesn't always require a massive cloud infrastructure.

Status: Highly Optimized

Users: 98 Active

Type: Voice Assistant

* Built with Python *

Honors & Awards

> Google Agentic League

Two-Time Selectee out of 1,200+ global applicants. Recognized for excellence in generative models & cutting-edge agentic architecture.

NYAS Junior Academy '26

Selected into the New York Academy of Sciences' Junior Academy, a highly selective global STEM program connecting young researchers with scientists to solve real-world challenges.

2nd Place – TeamsCode '26

#2 overall rank globally (1,160/1,200 pts). Competed entirely solo in a grueling 3-hour algorithmic programming contest.

Codeforces Expert

Elite competitive programming skills. High speed algorithmic problem-solving under extreme pressure.

StandOut Scholar

Merit scholarship for exceptional academic & extracurricular achievements in a highly selective (~6-15%) virtual internship.

International Computer Science Competition

Advanced to Finalist in the rigorous global qualification rounds.

1st Place National Coding Comp

Perfect score in Ascent Quest.

PromptWars '26

Selected for India's largest AI hackathon by Hack2skill.

The Knowledge Society

Admitted into a globally renowned (~15% acceptance) innovation program. Trained alongside top global youth.

Journey

Experience

  • Research Intern
    NIT Warangal Current

    Conducting applied research at the National Institute of Technology Warangal, working alongside faculty on technical problem-solving and analytical methods.

  • Student Researcher
    Science Mentorship Institute

    Engaged in hands-on computer science research under expert mentorship, building analytical capabilities and collaborating with peers.

  • Contributor
    GirlScript Summer of Code

    Collaborated with mentors and a global developer community to enhance real-world open-source projects.

  • Citizen Science Researcher
    NASA

    Contributed localized data for Aerosol Optical Depth and Cloud Cover to validate NASA satellite missions.

  • Public Developer
    Apple

    > Contributed as a Public Developer for iOS 16 and iOS 15.

Early Learning

Stanford

Code in Place '26

Selected for Stanford's CS106A intro program (~10% acceptance). Mastered foundational Python & problem-solving. Secured all badges.

Origin Story

Started coding at 13. SoloLearn Python (10+ certs).

Built Apple & Sony clone websites constantly pushing HTML/CSS boundaries.

Let's build something.

Have a project in mind, a research idea, or just want to say hi? Drop me a message below.