---
title: "Ankit Haldar - Lead Data Scientist in Greater Bengaluru Area | Tal"
description: "Meet Ankit Haldar, Staff Data Scientist | 10+ Yrs ML/AI for Churn, Retention & Monetization | Scaled Predictive Models to 12M+ Users | A/B Testing, MLOps, CRM Analytics | Ex-Confluent · Tesco · Brillio based in Greater Bengaluru Area. AI stack includes claude code, kimi, and langchain."
image: "https://tal.club/meet/ankit-haldar/share-image"
---

# Ankit Haldar on Tal

## Headline

Staff Data Scientist \| 10\+ Yrs ML/AI for Churn, Retention & Monetization \| Scaled Predictive Models to 12M\+ Users \| A/B Testing, MLOps, CRM Analytics \| Ex\-Confluent · Tesco · Brillio

## About

I build machine learning systems that move business numbers — not just model metrics\. Over 10\+ years as a Data Scientist, I've architected end\-to\-end predictive platforms across telecom, subscription retail, and high\-volume consumer businesses\. My work sits at the intersection of applied ML, revenue strategy, and cross\-functional execution: I design models that get deployed, adopted, and drive measurable outcomes\. Some highlights from the work I'm most proud of: \- Built a multi\-model churn propensity platform \(XGBoost/CatBoost ensemble\) across a 12M\+ subscriber base that identified 85% of at\-risk customers, reduced voluntary attrition by 15%, and delivered $25M/year in enterprise savings\. \- Pioneered an organization's ML experimentation framework — feature store, model registry, drift monitoring — cutting deployment cycle time by 50% and establishing standards adopted by three additional data science teams\. \- Orchestrated a hyper\-personalized retention offer engine with Marketing, Network Engineering, and Product leadership that boosted acceptance rates by 35% and recovered $18M annually\. \- Led demand forecasting initiatives using ARIMA/SARIMA across SKU × Channel × Location combinations, improving merchandise procurement accuracy by 6%\. \- Diagnosed hidden profit leakages in Central European retail operations through driver\-model analysis, enabling targeted interventions that recovered £2M in profit\. \- Mentored junior data scientists in advanced SQL, feature engineering, stakeholder communication, and ML best practices — building capability, not just models\. My toolkit spans Python, SQL \(BigQuery, Snowflake, Hive\), XGBoost, CatBoost, Apache Airflow, Tableau, GCP, and end\-to\-end MLOps infrastructure\. I'm hands\-on with code and strategy equally comfortable debugging a feature pipeline or presenting to C\-suite\. What drives me: translating complex analytical findings into decisions that non\-technical stakeholders can act on\. I believe the hardest part of data science isn't the model — it's the adoption\. I'm currently exploring Staff Data Scientist roles where I can define technical direction, scale ML platforms, and lead cross\-functional AI/ML strategy\. If your team is solving hard problems in churn, retention, personalization, or experimentation at scale, I'd love to connect\. Reach me at ankithaldar@gmail\.com or send a message here\.

## Location

Greater Bengaluru Area

## AI stack

claude code

kimi

langchain

langfuse

langgraph

langsmith

## Work experience

Lead Data Scientist at Brillio, Aug 2024 \- Present · 2 yrs 2 mos, Bengaluru, \- Churn Reduction: Developed a gradient\-boosted \(XGBoost, CatBoost\) propensity model identifying 85% of at\-risk customers \(top 3 deciles\), enabling targeted interventions that reduced annual churn by 15% \($25M saved annually\)\. \- ML Framework: Led end\-to\-end design of an ML experimentation pipeline \(feature store, model registry, monitoring\), cutting experimentation and deployment time by 50%\. \- Cross\-Functional Impact: Partnered with marketing to optimize a retention offer strategy, boosting acceptance rates by 35% and saving $18M annually\. \- Line\-Level Churn: Built a hierarchical model predicting churn at line/customer levels, improving accuracy by 18% vs\. legacy methods\. \- Customer Segmentation: Clustered high\-value subscribers using RFM \+ behavioral data, increasing upsell conversion by 27%\. \- Executive Reporting: Automated churn reporting in Tableau, reducing manual effort by 25\+ hours/month\. \- Campaign ROI: Analyzed 50\+ retention campaigns, identifying under\-performing segments and reallocating spend to save $2\.5M/year\. \- Root\-Cause Analysis: Mapped churn drivers \(network issues, billing complaints\) leading to a 15% reduction in service\-related cancellations\.

Lead Analyst – GTM Strategy and Planning at Confluent, Jan 2023 \- Apr 2024 · 1 yr 4 mos, Bengaluru, Karnataka, India, • Lead and mentor a team of 4 senior analysts and analysts\. • Designed, executed, monitored and statistically analyzed multiple split and multivariate A/B Tests increasing Product Qualified Leads \(PQLs\) by 17%\. • Collaborate with global leadership for design and implementation of comprehensive pipeline generation dashboard\. This interactive tool tracks individual and team performance, increasing 13% pipeline revenue, by assisting stakeholders with near real\-time insights to optimize sales strategies\. • Developed user driven, self\-service platform to track metrics for Quarterly Business Reporting to tracking progress in metrics daily over longer time\-frames \(daily, weekly, monthly\) with removal of all manual interventions\. • Partnered with the data engineering team to design and maintain new data pipelines in Apache Airflow to fetch previously missing or newly added attributes to Salesforce\. Maintained git repository to enhance created data pipelines for seamless reporting resulting in runtime reduction by 8%\. Tools: Tableau \| Python \| SQL\(BigQuery\) \| Apache\-Airflow \| Salesforce \| Git \| CRM Analytics

Lead Data Scientist at Tesco Bengaluru, Aug 2022 \- Jan 2023 · 6 mos, Bengaluru, Karnataka, India, • Led a team of 4 analysts to unveil previously unidentified profit leakages through a comprehensive driver model analysis\. This data\-driven approach pinpointed hidden loss factors within retail operations, enabling implementation of targeted mitigation strategies\. The resulting improvements yielded a combined profit increase of £2 million \(GBP\) across Central European stores\. • Spearheaded a new initiative focused on demand forecasting and demand sensing through ARIMA/SARIMA across SKU x Channel x Location combinations, factoring in promotions, weather patterns and industry events to predict future demand, improving Merchandise procurement by 6%\. • Natural Working Team \(NWT\) : Heading the CE team to help up\-skill Enterprise Analytics team by conducting various sessions and helping them with project roadblocks Tools: Tableau \| SQL \(Hadoop \| Hive \| Spark\) \| Python \(Pandas \| Numpy \| statmodels \| scikit\-learn \| seaborn\)

Senior Data Scientist at Tesco Bengaluru, Dec 2020 \- Jul 2022 · 1 yr 8 mos, Bengaluru, Karnataka, India, • Led team of analysts and category planners in collecting data and improving ways to measure KPIs to capture the most relevant and updated information\. • Developed an End\-2\-End automated Decision Cockpit for Central Europe execs and Category planners\. This cockpit keeps track of various crucial metrics to understand the health of the categories along with devising new strategies to improve the CPS \(Customer Priority Score\) of each category\. • Analyzed the Hypermarket de\-growth in CE region to reduce the loss due to Customer switching and increase of competition in various demographics\. The analysis is projected to reduce Customer switching by 2% in Czech Republic market\. • Create multiple automation in Python and VBA to enable automated reporting so as the decrease man hours\. Benefits of weekly 30\+ man hours were observed by the change\. Tools: Tableau \| SQL \(Hadoop \| Hive \| Spark\) \| Python \| Excel \| VBA

Senior Analyst at Affine Analytics, Jul 2019 \- Dec 2020 · 1 yr 6 mos, Bengaluru Area, India, Client: Mobile Game developer \(May 2020 \- Dec 2020\) ∗ Develop Apache\-Airflow based ETL process in Google Cloud Platform \(GCP\) for tracking in\-game metrics\. Analysed GCP usage and modified ETL process to reduce GCP costs by almost 50%\. ∗ Built highly customised Periscope dashboards for stakeholders to understand game performance\. ∗ Conceptualize new in\-game features and run A/B Tests\. Analyse A/B Test impact and suggest implementation of feature\. ∗ Undertake ad\-hoc investigations to understand player behaviour\. Discover new customer segments \(player clusters\) with monetization potential\. ∗ Implement new algorithm to track and eliminate/block cheaters from the game\. Tools: SQL \(BigQuery\) \| Apache\-Airflow \| Python \| Periscope \| Google Docs Suite \| Google Apps Script Client: Gaming Console Manufacturer \(Jul 2019 \- May 2020\) ∗ Implemented a regression model for forecasting of sales for upcoming major games with MAPE less than 15% and used Time series in forecasting sales of franchise games which further helped in deciding partnership with major game studios\. ∗ Apply regression modeling techniques to prioritize advertisement banners on store website to capture consumer attention boosting sales\. ∗ Implement model into product to simulate revenue based on advertisement priority\. ∗ Create reports on requests like used console and disk analysis, product churn and analyse performance and impact of promotions\. ∗ Analyse potential customer targeting strategies for upcoming titles by defining and quantifying different segments of potential audience\. ∗ Create Tableau dashboards for various business verticals that enabled clients to analyse business Tools: SQL \(Snowflake\) \| Python \| VBA \| Excel \| Tableau

Assistant Manager \- Analytics at WNS Global Services, Dec 2018 \- Jun 2019 · 7 mos, Bengaluru Area, India, • Offshore delivery & analytical lead \(4\+ members\) • Extensively used SQL/Python to mine and structure data from multiple loyalty & transactional tables to automate & derive KPIs for a 5M\+ loyalty customer base\. • Developed Tableau & Excel\-based dashboards to report Customer\-Beverage Affinity & Basket preferences, Store performance analysis, Loyalty Metrics Dashboard, Collaboration with third\-party services, etc\. • Evaluated potential drawbacks of Loyalty team’s collaboration with an insurance\- based third\-party service, which set up the grounds for renegotiating their deal, saving up to £2\.5M annually • Performed Look\-alike customer profiling and segmentation using RFM metrics and K means clustering, setting up a framework for Test vs Control Post Campaign/Event Analysis • Used ARIMA and Exponential Smoothing models to determine baseline sales & assess incremental impact of the launch of Red cups during the Holiday season \(UK market\) • Create fully automated data extractor & feeder by scraping data for franchises from multiple websites using python selenium\. Tools: SQL \(Oracle\) \| Python \| Tableau \| Excel \| VBA

## Education

Master of Business Administration \(M\.B\.A\.\), Business Analytics \(Conc\.\), Finance \(Spl\.\), Praxis Business School

Bachelor of Science \- BS, Information Technology, Kuvempu University, Shankaraghatta, Shimoga

[View the original profile](https://tal.club/meet/ankit-haldar)

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Person",
      "@id": "https://tal.club/meet/ankit-haldar#person",
      "name": "Ankit Haldar",
      "url": "https://tal.club/meet/ankit-haldar",
      "image": "https://tal.club/meet/ankit-haldar/profile-photo.jpg",
      "description": "Meet Ankit Haldar, Staff Data Scientist | 10+ Yrs ML/AI for Churn, Retention & Monetization | Scaled Predictive Models to 12M+ Users | A/B Testing, MLOps, CRM Analytics | Ex-Confluent · Tesco · Brillio based in Greater Bengaluru Area. AI stack includes claude code, kimi, and langchain.",
      "jobTitle": "Lead Data Scientist",
      "worksFor": {
        "@type": "Organization",
        "name": "Brillio"
      },
      "hasOccupation": {
        "@type": "Occupation",
        "name": "Lead Data Scientist"
      },
      "homeLocation": {
        "@type": "Place",
        "name": "Greater Bengaluru Area"
      },
      "alumniOf": [
        {
          "@type": "EducationalOrganization",
          "name": "Praxis Business School"
        },
        {
          "@type": "EducationalOrganization",
          "name": "Kuvempu University, Shankaraghatta, Shimoga"
        }
      ],
      "knowsAbout": [
        "claude code",
        "kimi",
        "langchain",
        "langfuse",
        "langgraph",
        "langsmith"
      ],
      "mainEntityOfPage": {
        "@id": "https://tal.club/meet/ankit-haldar#webpage"
      }
    },
    {
      "@type": "ProfilePage",
      "@id": "https://tal.club/meet/ankit-haldar#webpage",
      "url": "https://tal.club/meet/ankit-haldar",
      "name": "Ankit Haldar on Tal",
      "description": "Meet Ankit Haldar, Staff Data Scientist | 10+ Yrs ML/AI for Churn, Retention & Monetization | Scaled Predictive Models to 12M+ Users | A/B Testing, MLOps, CRM Analytics | Ex-Confluent · Tesco · Brillio based in Greater Bengaluru Area. AI stack includes claude code, kimi, and langchain.",
      "dateModified": "2026-09-01T09:44:58.031Z",
      "isPartOf": {
        "@id": "https://tal.club/#website"
      },
      "mainEntity": {
        "@id": "https://tal.club/meet/ankit-haldar#person"
      },
      "primaryImageOfPage": {
        "@type": "ImageObject",
        "url": "https://tal.club/meet/ankit-haldar/profile-photo.jpg",
        "contentUrl": "https://tal.club/meet/ankit-haldar/profile-photo.jpg",
        "caption": "Ankit Haldar profile photo on Tal"
      },
      "inLanguage": "en-IN"
    }
  ]
}
```