Naveen Kaushik

AI Researcher

West Delhi, Delhi, India10 yrs 7 mos experience

Highly StableAI Enabled

Key Highlights

Over a decade of experience in machine learning solutions.
Led impactful AI initiatives at Zeta.
Expertise in NLP and risk modeling.

Stackforce AI infers this person is a Data Scientist specializing in AI and machine learning for Fintech and Media industries.

Contact

naveenk903@gmail.com LinkedIn

Skills

Core Skills

Natural Language Processing (nlp)Risk ModelsGenerative AiMachine LearningRecommender SystemsData ScienceIos Development

Other Skills

Artificial Intelligence (AI)Artificial Neural NetworksBentoMLBig DataBig Data AnalyticsCC++CSSClusteringComputer VisionData AnalysisData ArchitectureDeep LearningDelegateEmail Prioritization Models

About

I am a Principal Data Scientist with over a decade of experience delivering scalable machine learning solutions across NLP, recommender systems, financial risk modeling, and conversational AI. At Zeta, I've spearheaded impactful AI initiatives, including building customer support bots for chat and voice channels, significantly reducing workload while maintaining high customer satisfaction. My technical leadership includes: Generative AI and LLM Development: Finetuning and deploying large language models tailored for specific business workflows, optimizing for latency and performance using frameworks like vLLM. Text-to-SQL Interfaces: Developed robust natural language query systems enabling analysts and operational teams to independently access data insights through auto-generated SQL queries. Risk Scoring Models: Led the development of credit risk models for loan origination, achieving substantial improvements in predictive performance and reduced risk exposure. Advanced Recommendation Systems: Created highly personalized content delivery and notification systems at Inshorts, significantly boosting user engagement through hybrid recommendation algorithms. I’m passionate about translating complex data-driven challenges into practical, impactful AI solutions. If you're exploring opportunities to harness machine learning and AI for strategic business outcomes, I'd be excited to connect.

Experience

10 yrs 7 mos

Total Experience

5 yrs 3 mos

Average Tenure

Current Experience

Zeta

2 roles

Principal Data Scientist II

Promoted

May 2024 – Sep 2025 · 1 yr 4 mos

Led the end-to-end development of chat and voice customer support bots, significantly reducing customer support workload without affecting customer satisfaction.
Finetuned Large Language Models (LLMs) using proprietary datasets and deployed efficiently with vLLM for high-throughput inference.
Developed a natural language interface enabling analysts and operations teams to automatically generate SQL queries, enhancing data accessibility and efficiency.
Designed and implemented XGBoost-based risk scoring models, achieving a 5% reduction in the bad rate compared to legacy models for loan origination.

Generative AILLM DevelopmentNatural Language Processing (NLP)Risk Models

Senior Data Scientist

Aug 2021 – Apr 2024 · 2 yrs 8 mos

Developed and deployed email prioritization models using BentoML, improving user engagement by ranking incoming emails by relevance.
Built merchant location estimation and anomaly detection models, enhancing the accuracy and effectiveness of transaction fraud detection systems.
Executed comprehensive credit card market segmentation studies aiding strategic program modeling and revenue forecasting.
Established MLFlow infrastructure for efficient experiment tracking across various ML projects.

Email Prioritization ModelsBentoMLFraud DetectionMarket SegmentationMLFlowMachine Learning

Inshorts

3 roles

Senior Data Scientist

Promoted

Apr 2021 – Jul 2021 · 3 mos · Noida, Uttar Pradesh, India

Enhanced personalized feed recommendation engines leveraging advanced hybrid models combining collaborative filtering and deep learning.
Improved summarization models through modified LexRank algorithms and transformer-based abstractive summarization techniques.

Personalized Feed RecommendationHybrid ModelsDeep LearningRecommender Systems

Data Scientist

Promoted

Sep 2015 – Mar 2021 · 5 yrs 6 mos · Noida, Uttar Pradesh, India

Developed multilingual full-text search systems, significantly enhancing content discoverability.
Built personalized push notification systems using clustering and logistic regression models, achieving approximately 40% higher click-through rates.
Created content and video recommendation engines employing embeddings and vector representations, improving user engagement on multiple platforms.

Multilingual Search SystemsPersonalized Push NotificationsClusteringLogistic RegressionData Science

iOS Developer

Jul 2014 – Aug 2015 · 1 yr 1 mo · Noida, Uttar Pradesh, India

Developed and maintained the foundational iOS app infrastructure, utilizing robust architectural patterns including MVC, Singleton, Delegate, and Notification.

iOS DevelopmentMVCSingletonDelegateNotification