Athlete of the Year
MIT Revels • 2022
A New Quest Begins
Neeraj
D Suvarna
About
Character Sheet
Neeraj
Explorer
Lv. 24
Hi, I'm Neeraj D Suvarna, a software engineer focused on building backend and AI-first products. I love shipping practical systems, turning problems into products, and constantly experimenting with new technologies.
MIT Manipal 2020-2024
Skills
PythonTypeScriptJavaScriptSQLFastAPIFlaskREST APIsWebSocketsPostgreSQLSupabaseQdrantLangGraphLlamaIndexWhisperMediaPipeAWSDockerRedis
Loves doodling · Loves travelling · Loves playing sports
Achievements Unlocked
Swimming Medalist
2022
Core Representative
MIT Athletics Team
Journey So Far
Experience
Every stop, a story. Every role, a relic.
Chapters Along The Trail
Moback Technologies India Pvt. Ltd.
Software Engineer
Mar 2025 - Present
Interview Coach
- Built an AI interview coach platform that generates personalized mock interview questions from the candidate's resume and job description.
- Developed the mock interview interface with multi-metric performance tracking, showing how the candidate's scores improve or decline across multiple interviews.
Multimodal RAG platform
- Built a multi-tenant RAG platform for creating organization-specific AI assistants over PDFs, images, videos, and internal knowledge sources.
- Developed async ingestion, vector retrieval, reranking, and citation-based responses so answers stay grounded in uploaded content.
PythonFastAPIFlaskPostgreSQLSupabaseLangGraphQdrantRAGWebSocketsWhisperMediaPipeAWSDocker
▸ Current Chapter
Moback Technologies India Pvt. Ltd.
Python Developer Intern
Sep 2024 - Feb 2025
Resume screening (ML)
- Built pipelines to score resumes against JDs with TF-IDF and BERT embeddings, annotated 100+ resumes and 50+ JDs in Label Studio.
- Improved screening precision by roughly 25–30%.
Pythonscikit-learnTF-IDFBERTLabel StudioNLP
MIT Manipal
Smart contract vulnerability detection (Ethereum)
Jan 2024 - May 2024
Final-year thesis
- Trained classical ML models (Random Forest, SVM, etc.) on Slither and AChecker features to detect access-control issues in Solidity contracts.
- ~87% accuracy with strong precision/F1 score, classical models beat RNN/TextCNN baselines on this dataset.
Pythonscikit-learnSlitherACheckerTF-IDFWord2VecFastText
Quest Log
Projects
Legendary encounters conquered, each worth remembering.
I
IC
Interview Coach
An AI interview practice platform that creates personalized mock questions from a resume and job description, then tracks performance across multiple interview attempts.
PythonFlaskPostgreSQLSupabaseSQLREST APIsWebSocketsWhisperPiper TTSMixpanel
II
RAG
Multimodal RAG Platform
A multi-tenant assistant platform for PDFs, images, videos, and internal knowledge sources, with async ingestion, retrieval, reranking, and grounded citations.
PythonFastAPILangGraphLlamaIndexQdrantRAGRedisDockerReact
III
RS
Resume Screening Model
An NLP and machine learning pipeline for matching candidate resumes with job descriptions using parsing, skill extraction, and role relevance scoring.
PythonNLPscikit-learnTF-IDFBERT EmbeddingsLabel Studio
IV
SC
Smart Contract Vulnerability Detection
ML models for detecting Ethereum smart contract access-control vulnerabilities using Slither and AChecker features, reaching about 87% accuracy.
PythonMachine Learningscikit-learnTF-IDFRandom ForestSVM
Treasure Found
Let's Connect
Take what you need, adventurer.