Amit Kumar, Developer in Dubai, United Arab Emirates
Amit is currently unavailable

Amit Kumar

AI Engineer and Developer

Dubai, United Arab Emirates

Toptal member since November 17, 2025

Bio

Amit is an AI engineer with 12 years of hands-on experience building production-grade generative AI and machine learning solutions, specializing in transforming complex business problems into scalable, high-impact AI products across healthcare, telecom, and insurance. He has led cross-functional teams and architected end-to-end platforms for global clients. With a focus on measurable outcomes, Amit has consistently delivered meaningful improvements in efficiency, automation, and decision-making.

Portfolio

M42
Python 3, Artificial Intelligence (AI), Natural Language Processing (NLP)...
Airtel India
Python 3, Apache Airflow, Speech-to-Text (STT), Text-to-Speech (TTS), Whisper...
Microsoft
Machine Learning, Artificial Intelligence (AI)...

Experience

  • Python 3 - 12 years
  • Pandas - 10 years
  • Natural Language Processing (NLP) - 10 years
  • Scikit-learn - 10 years
  • Machine Learning - 10 years
  • Large Language Model Operations (LLMOps) - 4 years
  • Large Language Models (LLMs) - 4 years
  • LangChain - 2 years

Preferred Environment

PyCharm, Visual Studio Code (VS Code), Python 3, LangChain, FastAPI, Asyncio, Data Science

The most amazing...

...solution I've developed is a web scraping utility that quickly gained traction, earning 84 GitHub stars and adoption across a global developer community.

Work Experience

Senior AI Engineer

2024 - PRESENT
M42
  • Built an AI-powered analytics platform that uses large language models and retrieval-augmented generation (RAG) to turn natural language queries into dynamic SQL and automated visualizations, enabling the analysis of clinical and environmental data.
  • Developed a physician-focused chatbot that facilitates access to patient EHR data by using a multi-agent framework and RAG to process information in FHIR format, improving efficiency and supporting better decision-making for healthcare providers.
  • Engineered an AI-assisted genetic variant pathogenicity classifier that searches and interprets scientific literature and bioinformatics data to assess pathogenicity in line with guidelines, reducing variant analysis time by four hours per patient.
Technologies: Python 3, Artificial Intelligence (AI), Natural Language Processing (NLP), Transformers, HL7 FHIR Standard, Electronic Health Records (EHR), Asyncio, FastAPI, LangChain, Visual Studio Code (VS Code), PyCharm, Retrieval-augmented Generation (RAG), Azure OpenAI Service, Python, Claude API, OpenAI API, OpenAI, Generative Artificial Intelligence (GenAI), Azure AI Studio, Claude, GraphDB, Amazon Bedrock, JSON, APIs, Architecture, Prompt Engineering, Product Development, NVIDIA CUDA, GPU Computing, PyTorch, Chatbots, AI Agents, LangGraph, NumPy, SQL, Back-end Architecture, CI/CD Pipelines, PostgreSQL, Document Parsing, Vector Search, Vector Databases, Data Science, IT Management, AI Programming, AI Voice Agents

Lead AI Engineer

2022 - 2024
Airtel India
  • Drove a team of six to develop a contact center call monitoring system powered by automatic speech recognition (ASR) technology.
  • Architected a lead-mining system for contact centers that analyzes 200,000 daily calls for potential leads, increasing lead conversion by 10%.
  • Built an intent classifier and an emerging-intent discovery model for a chatbot, increasing intent coverage by 16%.
Technologies: Python 3, Apache Airflow, Speech-to-Text (STT), Text-to-Speech (TTS), Whisper, NVIDIA NeMo, NVIDIA Triton, Asyncio, FastAPI, LangChain, Visual Studio Code (VS Code), PyCharm, Retrieval-augmented Generation (RAG), Python, Amazon Web Services (AWS), OpenAI API, OpenAI, Generative Artificial Intelligence (GenAI), JSON, APIs, Architecture, Prompt Engineering, Product Development, NVIDIA CUDA, GPU Computing, PyTorch, Chatbots, AI Agents, LangGraph, NumPy, SQL, Back-end Architecture, CI/CD Pipelines, PostgreSQL, Document Parsing, Vector Search, Vector Databases, Data Science, Business Intelligence (BI), IT Management, AI Programming, AI Voice Agents

Data and Applied Scientist II

2021 - 2022
Microsoft
  • Developed an aspect-based sentiment analysis system to extract insights and evaluate user sentiments from Windows-related social media data.
  • Engineered feedback-mining algorithms to analyze Feedback Hub data, identify high-impact bugs, and quantify the affected user base.
  • Implemented data pipelines to analyze telemetry and app usage data, revealing user behavior and informing feature prioritization.
Technologies: Machine Learning, Artificial Intelligence (AI), Natural Language Processing (NLP), Databricks, FastAPI, Visual Studio Code (VS Code), PyCharm, Python, JSON, APIs, Product Development, GPU Computing, PyTorch, NumPy, SQL, CI/CD Pipelines, PostgreSQL, Document Parsing, Vector Search, Vector Databases, Data Science, Business Intelligence (BI), IT Management, AI Programming

Tech Lead

2020 - 2021
InfoEdge
  • Spearheaded AI initiatives at Shiksha.com, establishing NLP and integration teams and building scalable compute infrastructure.
  • Directed the development of a virtual service agent, introducing new offerings for universities through Shiksha.com.
  • Developed a feedback analysis system for multi-dimensional college ratings, providing actionable insights for business decisions.
Technologies: Python 3, BERT, Transformers, Chatbots, Discovery, Feedback Review, Customer Feedback, FastAPI, PyCharm, Python, Google Cloud Platform (GCP), JSON, APIs, Product Development, GPU Computing, PyTorch, Keras, NumPy, SQL, Back-end Architecture, CI/CD Pipelines, PostgreSQL, Document Parsing, Vector Search, Data Science, IT Management, AI Programming

Senior AI Scientist

2018 - 2020
ExlService
  • Designed and deployed an AI-powered "Colleague Advisor" chatbot for the US Midwest region, enabling risk analysts to accelerate industry research and reducing risk report generation time by 10%.
  • Built a predictive Directors & Officers (D&O) liability model leveraging Bi-LSTM architectures and GloVe embeddings to assess litigation risk from legal text data.
  • Built a geospatial risk map integrating total insured value (TIV) with OpenStreetMap to help underwriters identify exposure concentrations and manage regional accumulation limits.
Technologies: Python, Flask API, Chatbots, OpenStreetMap, Churn Analysis, Amazon Lex, Amazon Web Services (AWS), NumPy, SQL, Back-end Architecture, CI/CD Pipelines, Vector Databases, Data Science, AI Programming

Experience

AI-powered Analytics Platform for Clinical Data

I developed an AI-powered analytics platform that allows users to query complex SQL databases using natural language. Leveraging advanced generative AI techniques, the system automatically generates SQL queries and visualization objects, delivering dynamic visualizations and actionable insights for fast, accurate, and intuitive data exploration.

Physician-focused AI Chatbot for EHR Access

I built a physician-focused chatbot that streamlines access to patient EHR data. By leveraging a multi-agent framework and RAG to process FHIR-formatted data, the chatbot improves efficiency and enables faster, more informed clinical decision-making.

Contact Center Call Monitoring Software

I developed an AI-powered contact center call monitoring system that securely analyzes advisor-customer interactions, automatically evaluates call quality, and identifies targeted training opportunities for advisors.

Emerging Intent Discovery for Chatbot

I architected an automated system to detect emerging user intents from conversational data using advanced NLP-based clustering techniques, improving intent coverage accuracy from 65% to 81% within one year.

Windows Feedback Analysis

I conducted text analysis on Windows 11 user feedback to identify problematic features and uncover underlying issues. I detected 150 critical and over 2,000 significant issues pre-release, enabling targeted resolution strategies and contributing to a smoother product launch.

Customer Churn Modeling

I developed a predictive system to identify customers likely to churn in the next renewal cycle, utilizing advanced feature engineering and data enhancement techniques to enhance the accuracy of future predictions.

Education

2009 - 2013

Bachelor's Degree in Engineering

Indian Institute of Technology Kanpur - Kanpur, India

Certifications

OCTOBER 2024 - PRESENT

Machine Learning in Production

Coursera

OCTOBER 2024 - PRESENT

Artificial Intelligence on Microsoft Azure

Microsoft

Skills

Libraries/APIs

Pandas, Scikit-learn, Matplotlib, NumPy, Claude API, OpenAI API, PyTorch, Asyncio, Keras, Natural Language Toolkit (NLTK), SpaCy, SciPy, Flask API

Tools

Azure OpenAI Service, Claude, PyCharm, Apache Airflow, Whisper, MATLAB, Gensim, Amazon Lex

Languages

Python 3, Python, C, SQL, Fortran

Frameworks

LangGraph

Storage

JSON, PostgreSQL

Paradigms

Model Context Protocol (MCP), Back-end Architecture, Business Intelligence (BI), HL7 FHIR Standard

Platforms

Azure, Amazon Web Services (AWS), Azure AI Studio, NVIDIA CUDA, Visual Studio Code (VS Code), NVIDIA NeMo, Databricks, Google Cloud Platform (GCP)

Other

LangChain, FastAPI, Large Language Models (LLMs), Large Language Model Operations (LLMOps), Artificial Intelligence (AI), Machine Learning, Natural Language Processing (NLP), AI Agents, Agentic AI, OpenAI, Generative Artificial Intelligence (GenAI), APIs, Architecture, Prompt Engineering, Product Development, GPU Computing, Document Parsing, Vector Search, Vector Databases, Data Science, AI Programming, Scientific Data Analysis, CI/CD Pipelines, IT Management, AI Voice Agents, Speech-to-Text (STT), Text-to-Speech (TTS), NVIDIA Triton, Machine Learning Operations (MLOps), BERT, Transformers, Chatbots, Discovery, Feedback Review, Customer Feedback, User Feedback, Electronic Health Records (EHR), Retrieval-augmented Generation (RAG), GraphDB, Amazon Bedrock, OpenStreetMap, Churn Analysis, Algorithms, Optimization, Engineering, Software

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring