---
title: "Umang Mittal - Software Engineer in Kota | Tal"
description: "Meet Umang Mittal, AI Engineer (S.E @MAQ Software) | LLM Systems · RAG · Agentic Pipelines | Ex-Cybersecurity | NSUT ’26 | Open to AI/FDE Roles based in Kota. AI stack includes hugging face, langchain, and langgraph."
image: "https://tal.club/meet/umang-13/share-image"
---

# Umang Mittal on Tal

## Headline

AI Engineer \(S\.E @MAQ Software\) \| LLM Systems · RAG · Agentic Pipelines \| Ex\-Cybersecurity \| NSUT ’26 \| Open to AI/FDE Roles

## About

I build AI systems — from LLM pipelines to agentic workflows — that solve real engineering problems\. Currently an SDE Intern at MAQ Software, where I work on production\-grade AI features used by enterprise clients\. Before this, I built AI solutions at Tech Swarm and did hands\-on penetration testing at CDAC — which means I think about both what systems can do and how they can break\. ───────────────────────────── WHAT I BUILD ───────────────────────────── → LLM\-powered pipelines \(RAG, embeddings, retrieval systems\) → Agentic workflows with tool use and memory → Data pipelines for ML\-ready datasets → Secure\-by\-default AI integrations ───────────────────────────── MY EDGE ───────────────────────────── Most AI engineers understand models\. I understand systems\. My background in cybersecurity means I build AI that's not just functional — it's robust\. I've gone from exploiting vulnerabilities to building the systems that need to resist them\. ───────────────────────────── CURRENT FOCUS ───────────────────────────── → Deepening expertise in multi\-agent systems & LLM deployment → Contributing to open\-source AI tooling → Actively exploring: AI Engineer / Applied AI / FDE roles at AI\-first companies post\-graduation \(July 2026\) ───────────────────────────── Let's connect if you're building with LLMs, working on frontier AI products, or looking for someone who ships\. ─────────────────────────────

## Location

Kota

## AI stack

hugging face

langchain

langgraph

openai

## Work experience

Software Engineer at MAQ Software, Jan 2026 \- Present · 7 mos, Noida, Uttar Pradesh, India, • Engineered an AI\-driven Azure DevOps management system for autonomous work item generation\. • Developed predictive machine learning models to forecast sprint delays and operational bottlenecks\. • Created a scalable multi\-agent automation system for 1,800 users, enhancing admin management efficiency\. • Automated large\-scale data scraping with the "EnergyAgentTool," ensuring high data integrity for operational intelligence\.

AI Intern at Tech Swarm, Jul 2025 \- Sep 2025 · 3 mos, Delhi, India, \- Built an end\-to\-end YouTube data extraction pipeline for B2B lead generation — ingesting product \+ company inputs, using an LLM to auto\-generate targeted search keywords, and scraping YouTube at scale to surface relevant video signals \- Engineered a Z\-score statistical filtering layer that eliminated 86% of low\-quality/irrelevant videos before any transcript processing — cutting LLM API costs and compute overhead dramatically while keeping the pipeline scalable \- Designed a two\-stage LLM processing system: first pass for keyword\-driven video discovery, second pass for deep transcript analysis to produce a standardized "relevance score" — transforming noisy, unstructured video data into a clean, machine\-readable signal \- Relevance scores were integrated directly into the company's lead scoring engine, improving B2B lead quality and accuracy for end customers

Penetration Tester at CDAC,Noida, Sep 2024 \- Oct 2024 · 2 mos

Penetration Tester at CDAC,Noida, Sep 2024 \- Oct 2024 · 2 mos

Logistics Coordinator at Moksha NSUT, Feb 2024 \- Mar 2024 · 2 mos, Delhi, India, \- Coordinated logistics for 50000\+ attendee college fest, managing 12 member volunteer team \- Streamlined vendor communication and scheduling, ensuring zero\-delay execution across 20 \+ events

Subject Matter Expert at PW \(PhysicsWallah\), Aug 2023 \- Mar 2024 · 8 mos

Student Mentor at PW \(PhysicsWallah\), Jun 2023 \- Aug 2023 · 3 mos, \- Created structured learning content for 1000\+ students across JEE courses on PW's platform \- Mentored 50\+ students \(Jun–Aug 2023\) through doubt resolution, contributing to 40% improvement in assessment scores \- Developed 10\+ practice modules and problem sets adopted into PW's official curriculum

## Education

Bachelor's degree, Computer Engineering, Netaji Subhas Institute of Technology

Schooling, Lawrence & Mayo Public School \- India

[View the original profile](https://tal.club/meet/umang-13)

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Person",
      "@id": "https://tal.club/meet/umang-13#person",
      "name": "Umang Mittal",
      "url": "https://tal.club/meet/umang-13",
      "image": "https://tal.club/meet/umang-13/profile-photo.jpg",
      "description": "Meet Umang Mittal, AI Engineer (S.E @MAQ Software) | LLM Systems · RAG · Agentic Pipelines | Ex-Cybersecurity | NSUT ’26 | Open to AI/FDE Roles based in Kota. AI stack includes hugging face, langchain, and langgraph.",
      "jobTitle": "Software Engineer",
      "worksFor": {
        "@type": "Organization",
        "name": "MAQ Software"
      },
      "hasOccupation": {
        "@type": "Occupation",
        "name": "Software Engineer"
      },
      "homeLocation": {
        "@type": "Place",
        "name": "Kota"
      },
      "alumniOf": [
        {
          "@type": "EducationalOrganization",
          "name": "Netaji Subhas Institute of Technology"
        },
        {
          "@type": "EducationalOrganization",
          "name": "Lawrence & Mayo Public School - India"
        }
      ],
      "knowsAbout": [
        "hugging face",
        "langchain",
        "langgraph",
        "openai"
      ],
      "mainEntityOfPage": {
        "@id": "https://tal.club/meet/umang-13#webpage"
      }
    },
    {
      "@type": "ProfilePage",
      "@id": "https://tal.club/meet/umang-13#webpage",
      "url": "https://tal.club/meet/umang-13",
      "name": "Umang Mittal on Tal",
      "description": "Meet Umang Mittal, AI Engineer (S.E @MAQ Software) | LLM Systems · RAG · Agentic Pipelines | Ex-Cybersecurity | NSUT ’26 | Open to AI/FDE Roles based in Kota. AI stack includes hugging face, langchain, and langgraph.",
      "dateModified": "2026-07-17T21:47:36.572Z",
      "isPartOf": {
        "@id": "https://tal.club/#website"
      },
      "mainEntity": {
        "@id": "https://tal.club/meet/umang-13#person"
      },
      "primaryImageOfPage": {
        "@type": "ImageObject",
        "url": "https://tal.club/meet/umang-13/profile-photo.jpg",
        "contentUrl": "https://tal.club/meet/umang-13/profile-photo.jpg",
        "caption": "Umang Mittal profile photo on Tal"
      },
      "inLanguage": "en-IN"
    }
  ]
}
```