---
title: "Jash Jummani - Associate Consultant - AI Engineer in Bengaluru | Tal"
description: "Meet Jash Jummani, Associate Consultant - AI Engineer @ KPMG | AI, Azure, FastAPI based in Bengaluru. AI stack includes claude code, hugging face, and langgraph."
image: "https://tal.club/meet/jash-jummani/share-image"
---

# Jash Jummani on Tal

## Headline

Associate Consultant \- AI Engineer @ KPMG \| AI, Azure, FastAPI

## About

Building AI systems that automate work so humans can spend less time fighting spreadsheets and more time solving real problems Passionate about building intelligent systems using LLMs, RAG, Agentic AI, NLP, and Deep Learning to solve real\-world problems\. Currently building scalable AI solutions and automation workflows at KPMG, with experience spanning multi\-agent systems, Computer Vision, NLP, and MLOps\. I enjoy turning complex ideas into practical AI products, exploring time\-series forecasting, and experimenting with intelligent agents that can actually get work done\. Published researcher in stock prediction using hybrid deep learning models, always curious about what’s next in AI\. Always open to connecting, collaborating, and talking AI \(or why models work perfectly… until demo day\)\. Python \| LangChain \| LangGraph \| PyTorch \| FastAPI \| Azure \| MLOps

## Location

Bengaluru

## AI stack

claude code

hugging face

langgraph

langsmith

mcp

githubcopilot

## Work experience

Associate Consultant \- AI Engineer at KPMG, May 2025 \- Present · 1 yr 6 mos, Bengaluru, Karnataka, India

Artificial Intelligence Engineer at KPMG, May 2025 \- Present · 1 yr 6 mos, Architected an ESG\-focused RAG application using Azure AI Search, Azure Blob Storage, and LangGraph for agent orchestration, enabling semantic retrieval and chat\-based insights that accelerated decision\-making\. Implemented a FastAPI\-based microservice to semantically index RFP documents and extract ESG\-relevant signals with 90%\+ labeling accuracy, resulting in a 40% reduction in analyst review effort\. Orchestrated an AI\-driven workflow for automated tax compliance validation, achieving a 65% decrease in manual checks and a 40% improvement in audit turnaround times\. Developed LangGraph agents using MCP for natural language control of Gmail and Calendar, facilitating schedule retrieval, meeting booking, email sending, and automatic categorization, which led to a 50% increase in task handling speed\.

Data Science Intern at QuantumStreet AI, Apr 2025 \- May 2025 · 2 mos, Remote, Evaluated the directional forecasting performance of state\-of\-the\-art time series models \(N\-HiTS, PatchTST, and Temporal Fusion Transformer\) across all Nifty\-50 stocks\. \- Built a directional evaluation framework measuring win rate, precision, recall, and F1\-score on up/down price movements across N\-HiTS, XGBoost, and ARIMA for all Nifty\-50 stocks\. \- Ran the analysis across the full Nifty\-50 universe to compare models and identify which perform best by stock and sector\. \- Analyzed results to surface patterns in where each model's directional signal is strongest and weakest\.

Data Scientist Intern at Eaton, May 2024 \- Jul 2024 · 3 mos, Pune District, Maharashtra, India, \-Developed an algorithm leveraging NLP techniques to extract and process information from PDFs for over 25 parameters\. \-Designed an infrastructure diagram with 94% precision by adhering to predefined rules and utilizing processed data\. \-Optimized the workflow to reduce diagram generation time by 40%, improving overall efficiency\.

Data Science Intern at Eaton, May 2024 \- Jul 2024 · 3 mos, Developed a PDF information extraction and processing algorithm for over 25 structured parameters using NLP techniques, entity recognition, and text classification\. Generated an Infrastructure Diagram with 94% precision from the gathered data, following predefined rules\. Achieved a 40% reduction in diagram generation time through effective optimization techniques\.

Teaching Assistant at Netaji Subhas University of Technology, Aug 2023 \- May 2025 · 1 yr 10 mos, New Delhi, Delhi, India, Part\-time Teaching Assistant for the M\.Tech \(Computer Science, AI\) program, supporting both core programming and applied machine learning coursework while contributing to faculty research\. Teaching \- Taught Data Structures and Algorithms, covering arrays, linked lists, stacks and queues, trees, graphs, sorting and searching, and dynamic programming, with an emphasis on problem solving and complexity analysis\. \- Supported Machine Learning coursework by helping students grasp core concepts, work through assignments, and debug model implementations\. \- Ran doubt\-clearing sessions and guided students in preparing for exams and coding interviews\. Research \- Partnered with a professor on research applying machine learning to stock price prediction, from data preparation and model experimentation to writing up results\. \- The work led to two peer\-reviewed conference publications: ICRTICC 2025 and AAIIC 2025\. Strengthened my ability to explain complex technical ideas clearly, a skill I use daily when working with cross\-functional teams in my current role\.

AI/ML Intern at LOGICWIND, Dec 2022 \- May 2023 · 6 mos, Surat, Gujarat, India, \-Engineered real\-time object detection projects utilizing 5\+ pre\-trained models, including YOLOv8, Detectron2, and ResNet\. \-Optimized algorithms and enhanced documentation, achieving a 200ms reduction in response time\.

## Education

Master of Technology \- MTech, Computer Science \(AI\), Netaji Subhas University of Technology

Computer engineering, Computer Engineering, Uka Tarsadia University

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

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Person",
      "@id": "https://tal.club/meet/jash-jummani#person",
      "name": "Jash Jummani",
      "url": "https://tal.club/meet/jash-jummani",
      "image": "https://tal.club/meet/jash-jummani/profile-photo.jpg",
      "description": "Meet Jash Jummani, Associate Consultant - AI Engineer @ KPMG | AI, Azure, FastAPI based in Bengaluru. AI stack includes claude code, hugging face, and langgraph.",
      "jobTitle": "Associate Consultant - AI Engineer",
      "worksFor": {
        "@type": "Organization",
        "name": "KPMG"
      },
      "hasOccupation": {
        "@type": "Occupation",
        "name": "Associate Consultant - AI Engineer"
      },
      "homeLocation": {
        "@type": "Place",
        "name": "Bengaluru"
      },
      "alumniOf": [
        {
          "@type": "EducationalOrganization",
          "name": "Netaji Subhas University of Technology"
        },
        {
          "@type": "EducationalOrganization",
          "name": "Uka Tarsadia University"
        }
      ],
      "knowsAbout": [
        "claude code",
        "hugging face",
        "langgraph",
        "langsmith",
        "mcp",
        "githubcopilot"
      ],
      "mainEntityOfPage": {
        "@id": "https://tal.club/meet/jash-jummani#webpage"
      }
    },
    {
      "@type": "ProfilePage",
      "@id": "https://tal.club/meet/jash-jummani#webpage",
      "url": "https://tal.club/meet/jash-jummani",
      "name": "Jash Jummani on Tal",
      "description": "Meet Jash Jummani, Associate Consultant - AI Engineer @ KPMG | AI, Azure, FastAPI based in Bengaluru. AI stack includes claude code, hugging face, and langgraph.",
      "dateModified": "2026-10-19T14:56:55.941Z",
      "isPartOf": {
        "@id": "https://tal.club/#website"
      },
      "mainEntity": {
        "@id": "https://tal.club/meet/jash-jummani#person"
      },
      "primaryImageOfPage": {
        "@type": "ImageObject",
        "url": "https://tal.club/meet/jash-jummani/profile-photo.jpg",
        "contentUrl": "https://tal.club/meet/jash-jummani/profile-photo.jpg",
        "caption": "Jash Jummani profile photo on Tal"
      },
      "inLanguage": "en-IN"
    }
  ]
}
```