resume

view résumé

education

Master of Science: Data Science and Artificial Intelligence
Saarbrücken, Germany2023 - Present
  • Natural Language Processing, Neural Networks, Generative AI, Machine Translation
  • Seminar: Machine Learning for NLP - challenges and applications of large language models
  • Seminar: Compositionality in Language & Computation - whether neural networks learn true compositional representations or merely memorize surface level patterns
Bachelor of Technology: Computer Science and Engineering
Mumbai, India2019 - 2023
  • Data Structures & Algorithms, Object Oriented Programming, Database Management Systems
  • Minor in Animation
  • Bachelor Thesis: Voice-to-Image Synthesis with NLP and Sentiment Analysis

work experience

InnoLab04.25 - Present

Werkstudent

Applied AI Development

  • Built asynchronous web scraping and entity extraction pipelines with bounded worker concurrency, processing 2,000+ pages across multiple domains.
  • Generated knowledge graphs from structured, semi-structured, and unstructured data using task-specific entity extraction with BAML and LangExtract, storing entities and relationships in SQLite when a dedicated graph database was unnecessary.
  • Parsed DOCX, Markdown, PDF, and PPTX documents for entity extraction, chunking, and storage using Docling and Unstructured, including content with tables, images, and multi-page layouts.
  • Evaluated LLMs by task fit, finding GPT-4o mini outperformed GPT-5 nano on tested tasks and selecting models based on output quality, efficiency, and cost.
  • Designed and experimented with agentic RAG workflows using LangChain and LangGraph, including sequential, supervisor, and adaptive agent architectures.
  • Built internal interactive demos using Gradio to prototype and evaluate LLM-driven workflows with stakeholders.
  • Investigated GraphRAG-based retrieval pipelines, conducting comparative experiments that informed the transition back to optimized vector-based RAG systems.
  • Implemented LangGraph-based evaluation pipelines assessing RAG outputs across multiple qualitative and retrieval-based criteria.
  • Integrated AI systems with OpenWebUI via custom API endpoints and supported switching between local and hosted LLM deployments.
  • Containerized and deployed internal demos using Docker for testing and experimentation on company infrastructure.
  • Improved retrieval performance through reranking strategies and prompt engineering for evaluation and multimodal captioning tasks.
  • Developed an IT support chatbot to guide customers through structured information gathering, reducing incomplete ticket submissions.
  • Built a dual-index RAG architecture (Knowledge Base + Solution Archive) with multilingual German/English support using ChromaDB and Jina Reranker.
  • Developed LangGraph pipelines for automated Jira ticket analysis, including priority classification, root cause extraction, and workaround identification with Pydantic-structured outputs.
  • Built a versioned prompt management system using markdown files for reproducible LLM workflows across evaluation and ingestion tasks.
Smart Service Engineering04.25 - 06.25

Student Assistant (HiWi)

Full Stack Web Development

  • Live website: iss.uni-saarland.de
  • Built a full-stack web platform using Django, enabling centralized management of team members, projects, publications, courses, and news content.
  • Implemented a role-based administration system supporting structured content creation and editing workflows.
  • Developed database backup mechanisms with scheduled and automated recovery support to ensure data reliability.
  • Delivered a production-ready system actively used by the research group for website and content management.
Government & Public Services03.23 - 09.23

Intern (Analyst)

Full Stack Development

  • Developed and integrated web application features using Spring Boot and deployed RESTful APIs for seamless communication.
PRISM Program09.21 - 02.22

Intern

Machine Learning & Data Privacy

  • Created a Personally Identifiable Information detection framework with 18 custom recognizers using NLTK and SpaCy, achieving 92% detection accuracy.

skills

Python • Kotlin • Java • TypeScript • JavaScript • SQL
Retrieval-Augmented Generation (RAG) • LangChain • LangGraph • RAG evaluation • LLM-as-judge evaluation • PyTorch • Hugging Face Transformers • Scikit-learn • Ragas • DeepEval • OpenAI API • Ollama • BAML • LangExtract • Knowledge Graphs • Langfuse • ChromaDB • Jina Reranker
Django • FastAPI • Spring Boot • RESTful APIs • Firebase
MySQL • PostgreSQL • SQLite • Neo4j • Firebase Realtime Database • SQLAlchemy
Next.js • React • Angular • Flask • Jinja2 • Gradio • Streamlit
Android • Jetpack Compose • Navigation Compose • DataStore • Readium • AndroidX PDF
Git • Docker • GitHub Actions • Cloudflare Workers • uv • Postman • Linux

projects

Creator & Android Engineer05.26 - Present

Independent Product Development

  • Designed and built Tether, a local-first Android reader for articles, RSS and Atom feeds, pasted text, PDFs, and EPUBs, with synchronized progress across RSVP and full-document reading.
  • Implemented the app in Kotlin and Jetpack Compose, using Readium and AndroidX PDF for retained-original publication reading, a durable location model for progress and annotations, and local storage for the library, settings, reading history, and backups.
  • Built a bounded Cloudflare Worker extraction fallback and an automated GitHub Actions release pipeline producing signed APK and Play-ready AAB artifacts; shipped v1.1 to Google Play closed testing without accounts, ads, or analytics.
  • Product details: tendril.works/tether
  • ProofRAG — Open-source Python CLI, Agent Skill, and GitHub Action for generating corpus-grounded RAG golden sets, running app predictions, evaluating retrieval and answer quality with LLM-as-judge metrics, producing HTML scorecards, and gating CI regressions.
  • rag-utils — Practical RAG script collection for DOCX/PDF processing, chunking, retrieval evaluation, context packing, query expansion, span tracing, SQLite embedding caching, semantic deduplication, and hybrid retrieval with reciprocal rank fusion.
  • Prescriptive Bias in LLM Sampling — Exploratory experiments measuring whether LLM samples drift from descriptive averages toward implicit ideal values across prompts, languages, temperatures, personas, and external baselines.
  • Program Repair and Hint Generation — Evaluated GPT-4o-mini and Phi-3-mini for Python program repair and educational hint generation on INTROPYNUS, using multi-candidate sampling, prompt engineering, LoRA fine-tuning, and multi-task learning.
  • Data Selection and PEFT — Combined influence-based data selection with BitFit, LoRA, and (IA)3 for MolFormer fine-tuning on molecular property prediction.
  • FactCheckLIAR — Hybrid BM25/FAISS fact-checking system with a fine-tuned BERT classifier, cached indexes, optional Ollama response generation, and Streamlit interface.
  • Image Segmentation with Scribble Supervision — Weakly supervised segmentation pipeline using scribbles to train image segmentation models with reduced annotation cost.
  • Healthcare Utilization Prediction — Predictive modeling workflow for healthcare utilization using structured clinical and demographic features.
  • VoxArt — Voice-to-image generation pipeline combining speech input, sentiment analysis, prompt construction, and image synthesis.

open source

  • microsoft/essex-toolkit — Added markdown code fence stripping before JSON parsing in _parse_json_string(), fixing JSONDecodeError failures caused by models (gpt-4o-mini, Ollama, LM Studio) that wrap structured outputs in fences. Fixes #437.
  • neo4j-labs/create-context-graph — Fixed connection lifecycle bug in reset_database() by wrapping Cypher execution in connect_neo4j() / close_neo4j() with try/finally cleanup. Fixes #26.
  • neo4j-labs/create-context-graph — Added --ontology-file CLI support for custom domain YAMLs, including ontology loading, validation, mutually exclusive domain options, generated data/ontology.yaml output, and regression coverage. Addresses #50.

languages

  • English: C1 (Fluent)
  • German: A2 (Learning)
  • Hindi: Native

achievements

  • 2024 Silver Medal for securing the 2nd rank in academics at Amity University.
  • 2024 Shri Baljit Shastri Award for exemplary human and traditional values.
  • 2022 Certificate of Excellence for outstanding performance at Samsung R&D Institute.