About
Data Scientist II with 6+ years of experience building and deploying predictive models across fintech, retail, and manufacturing domains. Proven track record in end-to-end ML lifecycle ownership — from data exploration and feature engineering to production deployment, monitoring, and recalibration. Experienced in credit risk-adjacent decision support systems, classification models with precision/recall tradeoff analysis, fraud-adjacent anomaly detection, and GenAI/LLM integration. Hands-on with AWS SageMaker, Azure ML, Docker, CI/CD pipelines, and real-time + batch scoring architectures. Comfortable operating with high autonomy and partnering directly with business stakeholders to translate risk into scalable, measurable ML systems.
Skills & Expertise (40)
Work Experience
Data Scientist
MathCo
Dec 2021 - Feb 2026
Led LLM integration for a GenAI-powered conversational chatbot on MathCo’s NucliOS platform, integrating Azure OpenAI models (GPT-3.5 and GPT-4o) to enable natural language querying across 9 datasets and 3 business domains, unlocking 120+ KPIs for self-serve analytics. Built using LangChain and RAG-based retrieval, the chatbot translated natural language to SQL, retained context across interactions, and achieved ~85% query accuracy with responses under 20 seconds. Built dedicated demand forecasting models for 26 automobile nameplates using clickstream behavioural data. Trained and benchmarked four models (XGBoost, LSTM, Gradient Boosting, and ARIMA) per nameplate and KPI, deploying the best-performing model to an automated AWS SageMaker pipeline — covering data ingestion, feature engineering, model training, drift monitoring, and output storage — to run monthly and generate next-month demand forecasts. Conducted a conjoint analysis for a beverage industry client to identify which product attributes (flavour, packaging, price point, size) would maximise sales for a new product launch. Built a Market Mix Model (MMM) using linear regression-based decomposition to isolate incremental sales contribution of four digital channels (Google Ads, Facebook, Instagram, Twitter) and calculate channel-level ROI. Developed a decision support tool for investor clients to evaluate whether portfolio companies met ESG compliance standards.
Machine Learning Developer
Triveni Global Software Technology Pvt. Ltd.
Jan 2019 - Jan 2021
Developed an intelligent Manufacturing Part Number (MPN) recommendation system by building a clustering model on unstructured part data to group similar MPNs. Deployed as a REST API using Azure ML, integrated into an existing web application with a custom UI, backend, and database connectivity — delivering a fully end-to-end production ML solution.
Education
M.Tech in Computer Science - Maharana Pratap University of Agriculture and Technology
2016 - 2019 · Afghanistan
Certifications
No certifications added yet
Interested in this developer?
Profile Score Breakdown
Profile Overview
Availability Details
Relocation
Open to Relocation
Skills (40)
Click a skill to find developers with the same skill
