---
title: "Surbhi Srivastava - AI Architect in Bengaluru | Tal"
description: "Meet Surbhi Srivastava, AI Architect based in Bengaluru. AI stack includes claude, hugging face, and langchain."
image: "https://tal.club/meet/surbhi-srivastava/share-image"
---

# Surbhi Srivastava on Tal

## Headline

AI Architect

## About

AI Architect and Data Science professional with 8\+ years of experience building scalable AI/ML solutions across financial crime analytics, regulatory compliance, risk modeling, NLP, and GenAI applications\. Skilled in designing end\-to\-end ML pipelines, anomaly detection systems, LLM\-powered NLP workflows, and production\-grade scoring platforms using Python, SQL, Databricks, Microsoft Fabric, AWS SageMaker, and MLFlow\. Experienced in deep learning, model validation, feature engineering, risk analytics, and governance frameworks, with a strong focus on translating complex analytics into business impact through stakeholder collaboration and scalable architecture design\.

## Location

Bengaluru

## AI stack

claude

hugging face

langchain

langgraph

langsmith

Cohere

## Work experience

AI Architect at Nagarro, Aug 2025 \- Present · 1 yr 2 mos, ● Developed a scalable Customer Satisfaction Risk\(CSR\) scoring platform on Databricks and Microsoft Fabric leveraging ML/DL capabilities for case\-level risk monitoring\. ● Built scalable feature engineering pipelines integrating structured data and LLM\-based NLP attributes, improving predictive signal quality by 15%\. ● Implemented feature evaluation frameworks covering exploratory analysis, selection logic, and correlation assessment to strengthen model robustness and feature reliability\. ● Created LLM\-based NLP workflows to process chronological texts, extract behavioral signals, translate multilingual content, and generate unstructured features for risk analysis, improving model performance by 20%\. ● Designed production\-grade batch scoring pipelines to generate daily CSR scores by orchestrating feature groups, model weights, and inference logic\. ● Defined solution architecture, pipeline documentation, feature store structures and MLFlow lifecycle integration in collaboration with cross\-functional teams to enable downstream analytics and long\-term maintainability\.

Manager: AI & Analytics at HSBC, Oct 2022 \- Aug 2025 · 2 yrs 11 mos, ● Designed and implemented Deep Learning based auto\-encoders for Anomaly Detection in trade transactions, reducing false positives by 30% and enhancing risk detection capability to safeguard the bank from regulatory censure and fines\. ● Conducted extensive exploratory data analysis \(EDA\) and feature engineering on large datasets, fine\-tuned models using advanced regularization and hyper\-parameters tuning, improving model convergence and performance\. ● Implemented advanced statistical techniques to compute anomaly thresholds, identifying high\-risk trades and reducing manual review time by 25%\. ● Established robust model monitoring frameworks, including data monitoring and performance tracking using KPIs such as model stability index, characteristic stability index, anomaly percentage, and reconstruction error, ensuring consistent model performance\. ● Contributed to advanced NLP\-driven communication surveillance capability, strategically enhancing bank's risk detection by 3 folds\. ● Performed model risk management responsibilities, encompassing model validation, inventory management, governance protocols, and reporting, ensuring compliance with multifaceted regulatory requirements and internal audit standards\. ● Led end to end experimentation life\-cycle: hypothesis design, roll\-out strategy, statistical analysis and impact assessment using tools like SQL, Python and statistical libraries\.

Business Intelligence Analyst at Empower, May 2020 \- Sep 2022 · 2 yrs 5 mos, ● Worked on advanced credit risk modeling solutions, leveraging machine learning algorithms and statistical analysis to strengthen business decision and risk assessment\. ● Implemented comprehensive data preparation pipelines, with domain specific feature engineering and extraction techniques to prepare high quality data resulted in enhanced model performance by 15%\. ● Worked on an unsupervised customer segmentation model to optimize financial plan recommendations, improving personalization and decision\-making\. ● Contributed in NLP\-based model to analyze customer\-raised event’s sentiment by implementing pipeline for test cleaning, preprocessing, word embedding to extract deeper insights\. ● Detected trend anomalies and outliers using statistical profiling and presented insights to business stakeholders through visualization and reports\.

Software Engineer \- ML at ZYCUS INFOTECH, Jul 2017 \- Apr 2020 · 2 y 9 mo, India, ● Worked on supplier performance analysis using SQL,classical machine learning, advanced analytics and provided insights using visualization tools\. ● Conducted customer analytics to derive actionable insights, improved customer engagement, and optimized business strategies\.

## Education

Bachelor of Technology, Computational Science, biju patnaik university of technology

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

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Person",
      "@id": "https://tal.club/meet/surbhi-srivastava#person",
      "name": "Surbhi Srivastava",
      "url": "https://tal.club/meet/surbhi-srivastava",
      "image": "https://tal.club/meet/surbhi-srivastava/profile-photo.jpg",
      "description": "Meet Surbhi Srivastava, AI Architect based in Bengaluru. AI stack includes claude, hugging face, and langchain.",
      "jobTitle": "AI Architect",
      "worksFor": {
        "@type": "Organization",
        "name": "Nagarro"
      },
      "hasOccupation": {
        "@type": "Occupation",
        "name": "AI Architect"
      },
      "homeLocation": {
        "@type": "Place",
        "name": "Bengaluru"
      },
      "alumniOf": [
        {
          "@type": "EducationalOrganization",
          "name": "biju patnaik university of technology"
        }
      ],
      "knowsAbout": [
        "claude",
        "hugging face",
        "langchain",
        "langgraph",
        "langsmith",
        "Cohere"
      ],
      "mainEntityOfPage": {
        "@id": "https://tal.club/meet/surbhi-srivastava#webpage"
      }
    },
    {
      "@type": "ProfilePage",
      "@id": "https://tal.club/meet/surbhi-srivastava#webpage",
      "url": "https://tal.club/meet/surbhi-srivastava",
      "name": "Surbhi Srivastava on Tal",
      "description": "Meet Surbhi Srivastava, AI Architect based in Bengaluru. AI stack includes claude, hugging face, and langchain.",
      "dateModified": "2026-09-30T16:29:01.420Z",
      "isPartOf": {
        "@id": "https://tal.club/#website"
      },
      "mainEntity": {
        "@id": "https://tal.club/meet/surbhi-srivastava#person"
      },
      "primaryImageOfPage": {
        "@type": "ImageObject",
        "url": "https://tal.club/meet/surbhi-srivastava/profile-photo.jpg",
        "contentUrl": "https://tal.club/meet/surbhi-srivastava/profile-photo.jpg",
        "caption": "Surbhi Srivastava profile photo on Tal"
      },
      "inLanguage": "en-IN"
    }
  ]
}
```