Yashsvi Dixit

Data Scientist

Noida, Uttar Pradesh, India6 yrs 8 mos experience

Most Likely To SwitchHighly Stable

Key Highlights

Increased user engagement by 20% through innovative algorithms.
Automated P&L processes, saving 200+ hours weekly.
Expertise in LLMs and market risk analysis.

Stackforce AI infers this person is a Data Scientist with expertise in Fintech and B2C applications.

Contact

Skills

Core Skills

Machine LearningLarge Language Models (llm)Market Risk Analysis

Other Skills

Deep LearningNatural Language Processing (NLP)Recommender SystemsCross-team CollaborationValue-at-Risk (VAR) CalculationsAutomationData AnalysisArtificial Intelligence (AI)Big DataA/B TestingSoftware DevelopmentApplied ProbabilityCommunicationStatisticsHypothesis Testing

About

ML/DL, GenAI, LLMs, Probability and Statistics, Problem Solving, Computer Science IIT Kharagpur

Experience

Inshorts

Data Scientist

Aug 2021 – Present · 4 yrs 7 mos

Responsible for Research and Development, Testing, and Scaling of Recommendation Algorithms for personalised news feed in the Inshorts news app.
Developed, end to end, a novel hyper-personalised recommendation algorithm for user-level personalised news article recommendations. Achieved ≈20% increase in average time spent by users on Inshorts app.
Used LLama2 and GPT LLMs for news summarisation and vectorisation.
Used LLMs for Entity Extraction from news and to divide users into topic cluster buckets.
Mitigation of critical challenges in scaling the algorithms to run realtime for over 1 million daily active users on Inshorts app.
User segment specific fine tuning of algorithm's parameters for further performance optimisation.
Close collaboration with Backend team, App team and Data Engineers.

Machine LearningLarge Language Models (LLM)Deep LearningNatural Language Processing (NLP)Recommender SystemsCross-team Collaboration

Goldman sachs

2 roles

Analyst

Jun 2019 – Jul 2021 · 2 yrs 1 mo · Bengaluru, Karnataka

Development of models required for assessing the performance of VaR models (in relation to P&L) and
quantifying the deficiencies in VaR models.
Automation and Improvement of the efficiency and reliability of the existing processes.
Developed a scalable framework for the attribution of daily portfolio P&L to market risk factors and the generation of automated text commentary. Product being used daily across business units, saving 200+ hours weekly for the team and eliminating operational errors.

Machine LearningMarket Risk AnalysisValue-at-Risk (VAR) CalculationsAutomationData Analysis