Shiv Markam

Data Scientist

Gurugram, Haryana, India4 yrs 7 mos experience

Highly StableAI Enabled

Key Highlights

Over 5 years of experience in Data Science.
Developed scalable AI models for fraud detection.
Holds 9 patents in fraud detection technology.

Stackforce AI infers this person is a Data Scientist with expertise in Fintech and AI-driven solutions.

Contact

Skills

Core Skills

Machine LearningSearchData ScienceFraud DetectionBig DataWeb Development

Other Skills

API DevelopmentNumpyFlaskAIAnnoy VectorDBBi-LSTMDomain KnowledgeNeural NetworksSWAXGBoostData AnalysisJavaScriptReactVRWeb ApplicationsRecommender Systems

About

I am a Data Scientist having experience of over 5 years working in Search, Card Payments and Fraud Detection Space. As part of my previous work experience I have leveraged AI to build real world products like enhanced customer search experience, improving fraud detection models, churn prediction.

Experience

4 yrs 7 mos

Total Experience

2 yrs 3 mos

Average Tenure

Current Experience

Info edge india ltd

Data Scientist

Oct 2022 – Mar 2024 · 1 yr 5 mos · Noida, Uttar Pradesh, India

Ranking Models (Higher Precision Final Layer)
◦ Developed Scalable END to END final layer Ranking model API (Gunicorn & Flask) for one of the Segments
◦ Enhanced the model’s robustness and explainability using Shap values, PDP Analysis, and AI Verification
◦ Optimized model Latency using Numpy code, aerospike DB features caching, configuring resource uses
Scalable Taxonomy Enhancement using AI
◦ Used Annoy VectorDB for duplicate candidates generation, bringing down O(Nˆ2) time complexity to O(Nlog(N))
◦ Used Character+Token Level Bi-LSTM Siamese Network identifying abbreviations, Spelling Mistakes
◦ Solution helped in improving existing taxonomy by 6% with minimal manual labor, domain knowledge expertise

Machine LearningSearch

Mastercard

2 roles

Data Scientist II

Promoted

Apr 2021 – Sep 2022 · 1 yr 5 mos

Improved the robustness of Neural Network Based Fraud Models by integrating SWA
Created new features to help lift the performance of existing fraud models
Built XGBoost based HPO Framework for Rapid Training , Validation and Testing
Modelling Chargeback Behavioural Risk at the account level
9 filed patents across patent offices of India and USA

Data ScienceFraud Detection

Data Scientist I

Jun 2019 – Mar 2021 · 1 yr 9 mos

Modelling Account Level Attrition Risk for the issuers to proactively identify and prevent Attrition
Entity Resolution Solution to Identify and Aggregate unique merchants from Merchant Database
Rule Based Transactional Volume Report Validation Program to identify inconsistencies in the data

Machine LearningBig Data

Tesco bengaluru

Software Engineer

May 2018 – Jul 2018 · 2 mos · Bengaluru Area, India

Built webapp for retail store planning and visualisation using , ReactVR
Webapp ingestes the store plan and product details to create an immersive 3D visualisation of store
Drag and Drop Functionality for users to place or remove a product on the virtual store shelves from available poduct catalogue

JavaScriptReactVRWeb Development