Hello, I’m Ankita Upadhyay

I build AI that deserves trust.

I am an AI and ML student, ELIZA Scholar, and software engineer based in Darmstadt. I am drawn to the human side of intelligent systems: how they explain themselves, support better decisions, and genuinely help people.

From building accessible education software to exploring reliable multimodal AI for scientific writing, I bring equal parts curiosity, care, and practical engineering.

Curious by nature, practical by training.

My journey moves between software engineering, machine learning, research, and the human questions that make technology worth building.

Where I am now

Researcher in progress

I am pursuing an MS in Artificial Intelligence and Machine Learning at TU Darmstadt as an ELIZA Scholar, with a growing interest in trustworthy and explainable AI.

What I bring

A builder’s mindset

Before research, I spent over two years building accessible, real time educational software. That experience shapes how I think about rigor, usability, and impact.

Beyond technology

Creative, people oriented, and curious

I enjoy reading, writing poems and songs, singing, and painting. I value ideas that bring people together and leave a positive mark.

Questions I am exploring

I am interested in methods that make AI systems reliable, interpretable, and useful in high stakes knowledge work.

Current work

Multimodal Evaluation of Scientific Documents

In an early stage literature review with Dr. Yue Feng at TU Darmstadt, I am investigating how multimodal AI can assess factual consistency and semantic alignment across scientific text, figures, tables, captions, and citations. My goal is to contribute to feedback that is reliable, explainable, and useful for scientific writing.

The path that brought me here

Each chapter taught me something different about solving problems, building for people, and staying curious.

May 2026 to present

Multimodal Evaluation of Scientific Documents

Early stage literature review with Dr. Yue Feng, Assistant Professor, ELLIS Member and Co Advisor at UKP Lab, TU Darmstadt

  • Reviewing reliable multimodal AI approaches for evaluating scientific documents.
  • Investigating factual consistency and semantic alignment across text, figures, tables, captions, and citations.
  • Developing a conceptual architecture for cross modal alignment, hotspot detection, and rubric guided assessment.

Jul 2023 to Oct 2025

Software Engineer, Zeus Learning

Mumbai, India

  • Built real time collaborative learning features with React and Redux, including complex shared state synchronization.
  • Developed discussion, media, and adaptive math game experiences for education platforms.
  • Wrote Playwright end to end test suites and applied WCAG accessibility guidelines across the UI.

Feb 2023 to July 2023

AI Research Intern, Lab Systems

Mumbai, India

  • Built a computer vision pipeline with CRNN and OCR for vehicle number plate recognition.
  • Automated image and metadata extraction from PDF, PPTX, and DOCX files using Python.

Research I have shared

Published work from my early projects, including the first ideas that led me towards research.

01

Saathi: An AI Chatbot

Research paper, June 2023. IEEE National Student’s Conference on Innovation in Rural Development, SSGMCE.

Read publication
02

Crowdfunding for Engineering Startups Using Blockchain and Smart Contracts

Research paper, April 2022. International Journal of Research and Analytical Reviews.

Read publication
03

Face Mask Detection and Attendance System

Research paper, June 2021. International Research Journal of Engineering and Technology.

Read publication
04

Face Mask Detection and Attendance System

Survey paper, March 2021. International Journal of Scientific and Research Publications.

Read publication

Things I have built

Projects where I combined curiosity, engineering, and the wish to make technology more useful.

July 2022

Saathi Your AI Companion

AI driven mental health support platform using deep learning, NLP, Flask API, and React. It earned 4th rank at IIT Bombay RESCON.

PythonTensorFlowReactNLP

2023

Luxa E Commerce Platform

A modern React e commerce application designed around clean product browsing and a smooth, responsive user experience.

ReactJavaScriptCSSUI Design

2023

Bitcoin Price Prediction LSTM Forecasting

A deep learning project that models Bitcoin price trends using LSTM based time series forecasting.

PythonTensorFlowLSTMTime Series

2023

Sarcasm Detection Text Classification

A TensorFlow based NLP project that classifies sarcasm in text through sequence modeling and contextual embeddings.

PythonTensorFlowKerasNLP

2022

Text Summarizer NLTK and Flask

A web application that condenses user supplied text with NLP techniques and adjustable summary length.

PythonNLTKFlaskSumy

2022

Invisible OpenCV Effect

A real time computer vision project that replaces a selected cloth colour with the captured background.

PythonOpenCVNumPyComputer Vision

The academic foundation behind my work.

A progression from computer engineering into artificial intelligence and machine learning research.

MS Artificial Intelligence and Machine Learning

Technical University of Darmstadt, Germany

2025 to 2027

ELIZA Scholar, Konrad Zuse School of Excellence in Learning and Intelligent Systems. Focus: foundation models, deep learning, ML systems.

BE Computer Engineering

University of Mumbai, India

2019 to 2023 | CGPA 9.5 / 10

Bachelor thesis: Saathi, AI Companion for Emotion and Mental Disorder Detection using Deep Learning.

Tools I use to turn ideas into systems.

A balanced toolkit shaped by research, full stack engineering, data work, and product thinking.

Machine Learning
Deep Learning, NLP, CNN, LSTM, Seq2Seq, Multi task Learning, PyTorch, TensorFlow, Keras, Hugging Face, Scikit learn, RASA
Frontend
React, Redux, JavaScript, HTML, CSS, Bootstrap, SASS, Material UI, Django, Browser Extension APIs, Figma
Backend and Data
Flask, Node.js, REST APIs, C#, PHP, MySQL, SQL, R
Programming
Python, JavaScript, C, C++, C#, PHP, SQL, Bash, Java
Tools and Cloud
Git, Docker, AWS, GCP, Linux, Playwright, OpenCV, Jupyter, Keras OCR
Web3
Blockchain, Ethereum, Smart Contracts
Languages
English (C1), Hindi (native), German (A1, in progress)
C
C
C plus plus
C++
CSS
CSS
Bootstrap
Bootstrap
Django
Django
Git
Git
Bash
Bash
Figma
Figma
Blockchain
Blockchain
Ethereum
Ethereum

Let’s connect.

For research collaboration, HiWi opportunities, or simply a thoughtful conversation about AI and people.

Email Ankita