---
title: "Ankit Sharma - AI Engineer in Bengaluru | Tal"
description: "Meet Ankit Sharma, AI Engineer at IBM | Former Senior Data Science Engineer at Cisco | NLP | Knowledge Graphs | Scalable ML Systems based in Bengaluru. AI stack includes claude code, claude, and codex."
image: "https://tal.club/meet/ankit-sharma-20/share-image"
---

# Ankit Sharma on Tal

## Headline

AI Engineer at IBM \| Former Senior Data Science Engineer at Cisco \| NLP \| Knowledge Graphs \| Scalable ML Systems

## About

AI & ML Engineer \| Agentic AI \(LangGraph, PydanticAI\) \| LLMs \| RAG \| Knowledge Graphs \| Enterprise AI As a AI/ML Engineer, I architect and scale enterprise\-grade AI solutions across Agentic AI, Retrieval\-Augmented Generation \(RAG\), and Natural Language Processing \(NLP\)\. I specialize in fine\-tuning, quantizing, and deploying production Large Language Models \(LLMs\) paired with stateful, type\-safe agent orchestration\. 🔹 Core Capabilities & Tech Stack: Agentic Orchestration: Building stateful, graph\-based multi\-agent systems with LangGraph \(cyclic logic, human\-in\-the\-loop\) and PydanticAI \(type safety, structured outputs, zero\-hallucination guardrails\)\. Frameworks & Databases: PyTorch, TensorFlow, Hugging Face, spaCy, Vector DBs, Neo4j Knowledge Graphs, Docker, and Kubernetes\. Optimization & Fine\-Tuning: PEFT, QLoRA, and GPTQ model quantization for low\-latency, low\-cost execution on constrained hardware\. 🔹 Proven Enterprise Impact: IBM / Cisco: Engineered RAG and agentic modules that boosted productivity by 25–60%, reduced query latency from 5s to under 2s, and improved retrieval accuracy using 10,000\+ PDF knowledge graphs\. Automation & Cost Reduction: Developed RAG document QA pipelines cutting deployment costs by 80%, alongside synthetic data generators outperforming standard baseline models\. 🔹 Leadership & Team Enablement: I drive bidirectional Knowledge Transfer \(KT\)—ingesting legacy architectures while training teams on modern agentic patterns, agent evaluation protocols, and GenAI best practices\. Active contributor to open\-source, tech speaker, and mentor , also keep myself updated with the market Technology enhancements\.

## Location

Bengaluru

## AI stack

claude code

claude

codex

hugging face

langfuse

langgraph

## Work experience

AI Engineer at IBM, May 2025 \- Present · 1 yr 6 mos, Bengaluru, As a AI Engineer at IBM designing and deploying scalable, enterprise\-grade Agentic AI and ML solutions with Go\. Specialist in multi\-agent orchestration and type\-safe systems using LangGraph \(stateful/graph logic\) and PydanticAI \(structured outputs & guardrails\)\. Expert in end\-to\-end model lifecycles, NLP, and knowledge engineering\. Active technical leader driving bidirectional Knowledge Transfer \(KT\) and team mentorship\. Key Achievements & Responsibilities Agentic Multi\-Agent Systems: Deployed resilient, multi\-agent workflows using LangGraph for cyclic state management/human\-in\-the\-loop and PydanticAI for strict output validation and dynamic tool execution\. Enterprise Knowledge Graphs: Architected a knowledge graph framework linking unstructured documents with contextual relationships, boosting search efficiency by 40%\. High\-Throughput ML/NLP Pipelines: Built end\-to\-end document classification and entity extraction pipelines using Hugging Face and spaCy, significantly reducing inference latency and cost\. Workflow Automation: Embedded autonomous AI agents into business workflows cross\-functionally, cutting manual processing time by 60%\. Enterprise AI Services: Built compliant AI services using Python, TensorFlow/PyTorch, and IBM Watson APIs , Azure and Claude aligned with governance standards\. Team Leadership & Knowledge Transfer \(KT\): Lead bidirectional KT—delivering sessions on agentic design patterns while acquiring context on incoming architecture requirements\. Mentored junior engineers, established agent evaluation protocols, and authored best\-practice documentation\.

Artificial Intelligence Engineer at IBM, May 2025 \- Present · 1 yr 6 mos, Engineer the IBM webMethods Hybrid Integration \(IWHI\) Integration Agent, translating natural\-language business intent into governed, multi\-stage integration workflows for technical and non\-technical users\.

Senior Data Science Engineer at Cisco, Dec 2023 \- May 2025 · 1 yr 6 mos, Bengaluru, Karnataka, India, \- Chatbot Development: \- Developed chatbots utilizing Retrieval\-Augmented Generation \(RAG\) with Knowledge Graph databases and Vector databases\. \- Created a Text\-to\-SQL retriever for interacting with PostgreSQL custom databases using Natural Language Processing \(NLP\)\. Continuously added new features and functionalities in production to enhance the SQL retriever's productivity for various use cases\. \- Fine\-tuned chatbots with open source models to improve performance and accuracy\. \- AI Agent Framework \(Exploring LangGraph, CrewAI , Autogen Frame work\): \- Developed an AI Agents framework using LangGraph, collaborating with the Product team to introduce new features and enhance customer experience\. \- Integrated innovative functionalities to streamline user interactions and optimize service delivery\. \- Advanced Fine\-Tuning: \- Implemented various fine\-tuning approaches for models, researching and applying state\-of\-the\-art techniques\. Specialized in fine\-tuning open source models using methods such as Reinforcement Learning from Human Feedback \(RLHF\), Direct Preference Optimization \(DPO\), and Supervised Fine\-Tuning \(SFT\) with the TRL library\. \- Research and Innovation: \- Continuously explored and integrated new AI and machine learning techniques to stay at the forefront of technology advancements\. \- Conducted experiments to validate the effectiveness of novel algorithms and methodologies in real\-world applications\.

Senior Data Engineer at Cisco, Dec 2023 \- May 2025 · 1 yr 6 mos, Architected enterprise Retrieval\-Augmented Generation \(RAG\) systems combining Knowledge Graphs, vector databases, embeddings, reranking, and open\-source LLMs for grounded question answering\.

Gen AI Researcher LLM at Infosys , IIIT Bangalore, Jul 2023 \- Nov 2023 · 5 mos, Bangalore, \- Open Source Model Implementation: \- Implemented Retrieval\-Augmented Generation \(RAG\) with LangChain, Llama Index, and Haystack\. \- Developed a duplicate defect search solution using Vector DB\. \- Created a Test Case ID finder for jira dump with RAG using open source LLM on client datasets\. \- Automated test case generation from open source models using RAG on client data\. \- LLM Fine\-Tuning: \- Fine\-tuning LLMs, including PyTorch, safetensors models, AWQ, GGUF, and GPTQ quantized models using PEFT, QLora, and SFT Trainer , DPO , ORPO\. \- Delivered multiple Gen AI knowledge transfer sessions to various IVSS teams within Infosys\. \- Client Engagements: \- Worked with clients such as Kone, CITI, AMEX, BT, State Street, JP Morgan for different GEN AI use case , delivered a GEN AI solutions for them\.

Artificial Intelligence Researcher at Infosys / IIIT Bangalore, Jul 2022 \- Dec 2023 · 1 yr 6 mos, Delivered Generative AI and LLM solutions for enterprise clients across financial services and engineering, including KONE, CITI, AMEX, BT, State Street, and JP Morgan\.

Associate Researcher at International Institute of Information Technology Bangalore, Jul 2022 \- Jun 2023 · 1 yr, Banglore, \- Collaborated with Infosys Ltd\. as a Research Associate, focusing on Multi Table relational synthetic data generation using Gaussian Mixture models\. \- Engaged in various machine learning \(ML\) and data mining \(DM\) technologies\. \- Applied Natural Language Processing in multiple in machine learning projects for different clients\. \- Worked on fine\-tuning open source large language models \(LLM\) and implementing models like GGUF and GPTQ , AWQ into development\.

## Education

Master of Business Administration, Computer Software Engineering, OP Jindal University

Master's degree, Artificial Intelligence and Machine learning, Liverpool John Moores University

[View the original profile](https://tal.club/meet/ankit-sharma-20)

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Person",
      "@id": "https://tal.club/meet/ankit-sharma-20#person",
      "name": "Ankit Sharma",
      "url": "https://tal.club/meet/ankit-sharma-20",
      "image": "https://tal.club/meet/ankit-sharma-20/profile-photo.jpg",
      "description": "Meet Ankit Sharma, AI Engineer at IBM | Former Senior Data Science Engineer at Cisco | NLP | Knowledge Graphs | Scalable ML Systems based in Bengaluru. AI stack includes claude code, claude, and codex.",
      "jobTitle": "AI Engineer",
      "worksFor": {
        "@type": "Organization",
        "name": "IBM"
      },
      "hasOccupation": {
        "@type": "Occupation",
        "name": "AI Engineer"
      },
      "homeLocation": {
        "@type": "Place",
        "name": "Bengaluru"
      },
      "alumniOf": [
        {
          "@type": "EducationalOrganization",
          "name": "OP Jindal University"
        },
        {
          "@type": "EducationalOrganization",
          "name": "Liverpool John Moores University"
        }
      ],
      "knowsAbout": [
        "claude code",
        "claude",
        "codex",
        "hugging face",
        "langfuse",
        "langgraph"
      ],
      "mainEntityOfPage": {
        "@id": "https://tal.club/meet/ankit-sharma-20#webpage"
      }
    },
    {
      "@type": "ProfilePage",
      "@id": "https://tal.club/meet/ankit-sharma-20#webpage",
      "url": "https://tal.club/meet/ankit-sharma-20",
      "name": "Ankit Sharma on Tal",
      "description": "Meet Ankit Sharma, AI Engineer at IBM | Former Senior Data Science Engineer at Cisco | NLP | Knowledge Graphs | Scalable ML Systems based in Bengaluru. AI stack includes claude code, claude, and codex.",
      "dateModified": "2026-10-06T14:28:02.747Z",
      "isPartOf": {
        "@id": "https://tal.club/#website"
      },
      "mainEntity": {
        "@id": "https://tal.club/meet/ankit-sharma-20#person"
      },
      "primaryImageOfPage": {
        "@type": "ImageObject",
        "url": "https://tal.club/meet/ankit-sharma-20/profile-photo.jpg",
        "contentUrl": "https://tal.club/meet/ankit-sharma-20/profile-photo.jpg",
        "caption": "Ankit Sharma profile photo on Tal"
      },
      "inLanguage": "en-IN"
    }
  ]
}
```