---
title: "Muskan Das - Data Scientist in New York City Metropolitan Area | Tal"
description: "Meet Muskan Das, Data Scientist | ML‑Driven | Agentic AI Engineer | LLM Orchestration | $2M Impact | M.Sc. Statistics and Machine Learning | 150K predictions/day | 85% faster | 35% growth | Ex-Infosys based in New York City Metropolitan Area. AI stack includes antigravity, deepmind, and github copi..."
image: "https://tal.club/meet/muskan-das-1/share-image"
---

# Muskan Das on Tal

## Headline

Data Scientist \| ML‑Driven \| Agentic AI Engineer \| LLM Orchestration \| $2M Impact \| M\.Sc\. Statistics and Machine Learning \| 150K predictions/day \| 85% faster \| 35% growth \| Ex\-Infosys

## About

As a seasoned professional in the fields of data science and business analysis, I possess a strong passion for Artificial Intelligence and Natural Language Processing\. My expertise lies in formulating precise problem statements, conducting extensive data analysis, and developing robust models to derive valuable insights\. Throughout my career, I have collaborated with leading organizations to tackle complex data analytics challenges, consistently delivering impactful results that have influenced key decision\-making processes\. I thrive on utilizing my skills and experience to drive innovation and generate actionable solutions in the realms of data science and business analysis\.

## Location

New York City Metropolitan Area

## AI stack

antigravity

deepmind

github copilot

ChatGPT Images

Mistral Vibe

OpenAI Playground

## Work experience

Data Scientist at Spotline, Inc\., Sep 2023 \- Present · 3 yrs 2 mos, New York City Metropolitan Area, • Architected a production\-grade Financial AI Copilot with multi\-layer LLM orchestration \(SLMs/LLMs\), intent\-based query routing, schema\-aware Python data\-access classes, and strict SQL templates — eliminating mis\-routing and hallucination at scale • Engineered a Hybrid RAG pipeline on OpenSearch Neural Search \+ dense embeddings to surface historical failure patterns and financial impacts, cutting root\-cause analysis time by 85% and improving next\-quarter forecast accuracy by 18% • Integrated real\-time anomaly detection models \(Prophet, Isolation Forest, LSTM\) into a Kafka → Python ETL → PostgreSQL pipeline, enabling 150,000\+ daily turbine forecasts with EBITDA impact analysis delivered directly via chat • Deployed the copilot on AWS SageMaker GPU instances with CI/CD pipelines, supporting live production traffic from Suzlon’s finance and operations teams • Reduced MAPE from 8\.3% → 4\.9% on energy generation forecasts; enabled 18–24 hour early fault detection across wind turbine clusters Skills: LLM Orchestration · RAG · OpenSearch · LangChain · Prophet · LSTM · Isolation Forest · Kafka · PostgreSQL · AWS SageMaker · Python · FastAPI · Agentic A

Postgraduate Research Assistant at MIT App Inventor, Jan 2023 \- May 2023 · 5 mos, Troy, New York, United States, During my research at MIT, I had the opportunity to work on a fascinating project involving the integration of Raspberry Pi 4B with MIT App Inventor\. 1\. As part of this project, I conducted a comprehensive analysis and prepared a detailed report on the seamless integration and functionalities achieved\. 2\. In collaboration with a team of talented individuals, we explored the potential of combining MIT App Inventor's intuitive visual programming environment with the powerful capabilities of Raspberry Pi 4B\. 3\. I conducted extensive research on the technical aspects of integrating Raspberry Pi 4B with MIT App Inventor, integrating the methods of IOT, including hardware configuration, communication protocols, and software development\. I meticulously analyzed the compatibility, performance, and potential challenges associated with this integration\. 4\. I successfully implemented various functionalities and features by harnessing the capabilities of both MIT App Inventor and Raspberry Pi 4B\. This included utilizing GPIO pins, accessing sensors, controlling actuators, and establishing communication between the app and the Raspberry Pi board\. 5\. I documented my findings, insights, and recommendations in a comprehensive report\. The report highlights the benefits, limitations, and potential use cases of integrating Raspberry Pi 4B with MIT App Inventor, providing valuable insights for future projects and endeavors in the field of app development and Internet of Things \(IoT\)\. 5\. My experience with MIT App Inventor and Raspberry Pi 4B has enriched my technical as well as reinforced my passion for innovation\.

Machine Learning Intern at Amazon, Jan 2023 \- May 2023 · 5 mos, United States, • Built customer segmentation models \(KNN, K\-Means, Random Forest\) on Amazon Ads data to identify high\-value vendor cohorts, improving ad targeting accuracy by 20% and boosting campaign ROI by 15% • Designed predictive ad optimization models using Random Forest and K\-Means clustering; developed SQL queries to enhance generative AI context embedding accuracy for downstream recommendation pipelines • Conducted root\-cause analysis on ad performance data for investor reporting, uncovering behavioral trends across upstream vendors that drove 15% performance improvement in targeted markets • Built automated reporting dashboards \(Tableau/Plotly\) to surface KPIs, enabling stakeholders to draw forward\-looking inferences on customer growth trajectories and ad spend efficiency • Applied regression\-based sales forecasting \(SAS JMP\) with 12% accuracy improvement, supporting data\-driven budget allocation decisions for Amazon Ads campaigns Skills: Machine Learning · Customer Segmentation · Random Forest · K\-Means · SQL · Predictive Modeling · A/B Testing · Tableau · Python · Scikit\-learn · Business Analytics

Data Analyst at Infosys, Oct 2020 \- Aug 2022 · 1 yr 11 mos, Bengaluru, • Developed predictive risk analytics models \(XGBoost, Random Forest, K\-Means\) for 500\+ banking portfolios, driving a 35% increase in client acquisition and supporting $2M\+ in additional revenue • Automated time\-series forecasting and supervised learning pipelines for campaign strategy, regulatory reporting, and fraud detection; implemented PSI\-based drift detection, sustaining 99%\+ model accuracy in production • Delivered AUC improvement from 0\.61 → 0\.74 on portfolio risk classification models through feature engineering, hyperparameter tuning, and cross\-validation • Improved campaign ROI by 25% through client segmentation models and data\-driven targeting strategies deployed across wealth management workflows • Led UAT, change management, and stakeholder training for ML solution rollouts, achieving 20–25% efficiency gains across reporting and analytics teams Skills: XGBoost · Random Forest · Time Series · Fraud Detection · SQL · Regulatory Reporting · Model Monitoring · Drift Detection · Python · Tableau · Power BI

Intern at Infosys, Jan 2020 \- Sep 2020 · 9 mos, Mysore, • Automated data engineering tasks and deployed scalable ETL scripts for production, reducing manual reporting time by 30% and supporting real\-time analytics for 4,000\+ concurrent users

## Education

Master's degree, Business analytics, Rensselaer Polytechnic Institute

Bachelor of Technology \- BTech, Computer Science, Siksha 'O'​ Anusandhan University

## Public links

- github: https://github\.com/dasm3

[View the original profile](https://tal.club/meet/muskan-das-1)

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Person",
      "@id": "https://tal.club/meet/muskan-das-1#person",
      "name": "Muskan Das",
      "url": "https://tal.club/meet/muskan-das-1",
      "image": "https://tal.club/meet/muskan-das-1/profile-photo.jpg",
      "description": "Meet Muskan Das, Data Scientist | ML‑Driven | Agentic AI Engineer | LLM Orchestration | $2M Impact | M.Sc. Statistics and Machine Learning | 150K predictions/day | 85% faster | 35% growth | Ex-Infosys based in New York City Metropolitan Area. AI stack includes antigravity, deepmind, and github copi...",
      "jobTitle": "Data Scientist",
      "worksFor": {
        "@type": "Organization",
        "name": "Spotline, Inc."
      },
      "hasOccupation": {
        "@type": "Occupation",
        "name": "Data Scientist"
      },
      "homeLocation": {
        "@type": "Place",
        "name": "New York City Metropolitan Area"
      },
      "alumniOf": [
        {
          "@type": "EducationalOrganization",
          "name": "Rensselaer Polytechnic Institute"
        },
        {
          "@type": "EducationalOrganization",
          "name": "Siksha 'O'​ Anusandhan University"
        }
      ],
      "knowsAbout": [
        "antigravity",
        "deepmind",
        "github copilot",
        "ChatGPT Images",
        "Mistral Vibe",
        "OpenAI Playground"
      ],
      "sameAs": [
        "https://github.com/dasm3"
      ],
      "mainEntityOfPage": {
        "@id": "https://tal.club/meet/muskan-das-1#webpage"
      }
    },
    {
      "@type": "ProfilePage",
      "@id": "https://tal.club/meet/muskan-das-1#webpage",
      "url": "https://tal.club/meet/muskan-das-1",
      "name": "Muskan Das on Tal",
      "description": "Meet Muskan Das, Data Scientist | ML‑Driven | Agentic AI Engineer | LLM Orchestration | $2M Impact | M.Sc. Statistics and Machine Learning | 150K predictions/day | 85% faster | 35% growth | Ex-Infosys based in New York City Metropolitan Area. AI stack includes antigravity, deepmind, and github copi...",
      "dateModified": "2026-10-30T15:40:58.565Z",
      "isPartOf": {
        "@id": "https://tal.club/#website"
      },
      "mainEntity": {
        "@id": "https://tal.club/meet/muskan-das-1#person"
      },
      "primaryImageOfPage": {
        "@type": "ImageObject",
        "url": "https://tal.club/meet/muskan-das-1/profile-photo.jpg",
        "contentUrl": "https://tal.club/meet/muskan-das-1/profile-photo.jpg",
        "caption": "Muskan Das profile photo on Tal"
      },
      "inLanguage": "en-IN"
    }
  ]
}
```