Naman Gupta

AI Researcher

Bengaluru, Karnataka, India7 yrs 7 mos experience
Most Likely To SwitchAI Enabled

Key Highlights

  • Lead Data Scientist with 2+ patents in AI.
  • Expert in Generative AI and Machine Learning.
  • Strong background in Quantitative Finance and Trading.
Stackforce AI infers this person is a Data Science expert with a focus on Generative AI and Quantitative Finance.

Contact

LinkedIn

Skills

Core Skills

Data ScienceMachine Learning

Other Skills

Natural Language Processing (NLP)Retrieval-Augmented Generation (RAG)Reinforcement LearningQuantitative AnalyticsGenerative AI ToolsPySparkElastic SearchFlaskKnowledge GraphsRecommender SystemsMachine Learning AlgorithmsBig DataRecommendation SystemsRegression AnalysisTopic Modeling

Experience

7 yrs 7 mos
Total Experience
2 yrs 8 mos
Average Tenure
6 yrs 1 mo
Current Experience

Jio

Lead Data Scientist

Aug 2020 – Present · 5 yrs 10 mos · Bengaluru, Karnataka, India

  • Versa Chat Assistant
  • A GenAI powered multimodal and multilingual chat assistant to automate customer 360 profile across all Jio Retail Platforms (Ajio, Reliance Digital, Netmeds, Jiomart)
  • Jio Indic Lens
  • Indic Search and Discovery platforms which supports billions of SKUs under one hood. This solution helps discover hidden insights applying GenAI and AI Search solutions to revolutionize customer search and purchase journey. (This technology has been patented by Jio Platforms.)
  • Jio Search and Discovery Platform
  • Deployed highly scalable ML pipeline (PySpark) using Elastic Search, Flask and Redis for Search Engine Auto-Correction and Auto-Suggestion tasks used for Auto Spell Correction, Auto Suggestion, Hierarchial Ranking of results. Business impact : improvement of recall rate to 0.94 and CTR score of 0.44 !
  • Jio RecoAI
  • User Recommender system for JioMart powering brands, products, all product level category substitutes for Knowledge graph entity recognition.
  • Jio Ashtang Project (Jio Brain)
  • Extracted growth rules using In-house formulation (combining Lift score, Log-odd ratio, Confidence score) for Personalised Product Feed Sugesstion Pipeline.
  • ML Research Projects @JIO -
  • Model Retrofitting Pipeline
  • Building SOTA model for Twitter Hatespeech, SemEval Dataset using model retrofitting techniques. Improved existing model accuracy by 6% .
  • Indic Character Transition and Phonetic Variants using Gaussian Kernel Estimators
  • Character correction in spell variants across all indic dialects (hindi, marathi, bengali , etc.)
Natural Language Processing (NLP)Retrieval-Augmented Generation (RAG)Reinforcement LearningQuantitative AnalyticsData ScienceMachine Learning

Self-employed

Quant Researcher - Equity and Derivatives

May 2020 – Present · 6 yrs 1 mo

Dailyhunt (versé innovation pvt ltd)

2 roles

Data Scientist

Aug 2019 – Aug 2020 · 1 yr

  • Honing Distributed Technology skills with Big Data (Pyspark) and applying Machine learning algorithms at scale for production.
  • News-Entity Knowledge Graph(TAU)
  • Developed News-Entity Knowledge Graph with entity nodes (such as NER, Category, Interest Groups, etc.), generated by association rules on user interests and their browsing history and trending NERs.
  • User-Entity Recommendation
  • Built User Recommendation Pipeline, using Geo Co-Located User Service, News-Entity Knowledge Graph and Marcov Chain Models based on User Events(Card-Clicks , Share , Like , etc).
  • Publisher Sensitivity Model
  • Analysed Publisher Data using Regression approach to rank and segregate spurious content (by highly bounced publishers) from magnanimous content (by highly sensitive publishers).

Software Engineer Intern

Feb 2019 – Jul 2019 · 5 mos

  • LDA Topic Modelling
  • Clustering on Publisher Articles (documents) to extract top words (using SVD and prob. distribution matrices) and bucketing top words into 128 to 512 categories to classify and tag latest published articles into various categories (Sports, Politics, Finance, etc).

Jaypee institute of information technology, noida

Undergraduate Research Assistant

Nov 2018 – Jan 2019 · 2 mos · Noida Area, India

  • Speaker Recognition System ( IEEE Publication )
  • Research Work on Speaker Voice verification using Siamese Network (trained on persons' voice mel-spectrogram data on VGG-16 Model).

Pucho technology information private limited

Deep Learning Intern

Aug 2018 – Dec 2018 · 4 mos

  • Working on Speaker recognition and speaker authentication .

Jiit , mathematica hub

Coordinator

Jul 2017 – Dec 2017 · 5 mos · Noida

Education

WorldQuant University

Master of Science - MS

Jul 2022 – Jul 2024

Jaypee Institute of Information Technology, Noida

Bachelor of Technology — Computer Science

Jan 2015 – Jan 2019

WorldQuant University

Financial Engineering — Econometrics and Quantitative Economics