Vinayak Mathur profile photo 1 on Tal

Vinayak Mathur

AI Engineer | Expert in Generative AI & Cloud Solutions | Nvidia Certified Generative AI Associate | Deloitte | ex-Cognizant | 4+ years of Experience

Hyderabad

5+ years of work experience

Deloitte
Nov 20251 yr 1 moPresent

Hyderabad

Cognizant
Jul 20254 mosOct 2025

Gurugram, Haryana, India

- Led the architecture and end-to-end development of a generative AI-powered data analyst for a global beverage leader. The agent automates BI workflows by translating user questions into SQL queries, data visualizations, and full analytical reports. - Engineered the system's core intelligence through a sophisticated prompt library, defining complex business rules for YTD analysis and market share calculations. - Built a resilient, multi-agent system (using LangGraph) where a central orchestrator delegates tasks to specialized agents for SQL generation, data visualization, and advanced Pandas-based calculations. - Integrated key Azure services, including Azure OpenAI for reasoning, Azure Blob Storage for stateful conversation persistence, and Azure SQL for live data querying.
Cognizant
Jul 20223 yrs 1 moJul 2025

Gurugram, Haryana, India

Multi-Agent Autonomous System Framework (Nvidia NIM, RAG, Python) • Architected and implemented a fully autonomous multi-agent system with Nvidia NIM, orchestrating specialized agents across diverse enterprise tools. • Developed a Linux system control agent for complex command execution without human intervention, and an enterprise communication agent handling Outlook/Teams interactions via Microsoft Graph API. • Engineered a development environment agent for hardware/software configuration, integrated a RAG pipeline for context-aware debugging, and set up agent-to-agent communication protocols for collaborative problem-solving. Automated Interview Platform (Generative AI, Kubernetes, Azure, AWS, GCP) • Created a full-stack automated interview solution leveraging LLMs (OpenAI GPT-4, Anthropic Claude), incorporating real-time speech processing. • Deployed Speech-to-Text and Text-to-Speech services on Kubernetes, achieving 99.9% uptime while reducing costs by 40%. • Implemented a multi-cloud strategy (Azure, AWS, GCP) with a unified CI/CD pipeline to streamline updates and resource allocation. Big 4 Financial Audit Automation & Analysis System (Generative AI, Python, Azure) • Developed a RAG-powered audit workflow for a Big 4 firm, automating lead sheet creation and analysis with GPT-3 and GPT-4. • Employed Python libraries (Openpyxl, Camelot, pdfPlumber) to parse financial data from PDFs, maintaining table integrity. • Built a user-friendly Streamlit interface that cut audit prep time by 65% and increased accuracy by 30%. Advanced Research Projects • Led comparative analysis of emerging LLMs (GPT, Claude, Llama, Gemini, Deepseek) to assess enterprise application viability. • Explored knowledge graph techniques for semantic relationship extraction from unstructured data. • Investigated prompt engineering, data generation, and system enhancements to optimize AI performance and reliability.
Cognizant
Jan 20227 mosJul 2022

- Specialized in AI/ML and worked on multiple projects involving AIML, NLP, and data analytics during GENC Next Internship at Cognizant. - Expanded skills in Pandas, Time Series Analysis, Statistical Data Analysis, AWS SageMaker, Amazon EC2, Apache Spark, Microsoft Azure, SQL, and Machine Learning through hands-on experience and training sessions. - Developed a strong foundation in data analysis and AI technologies, which helped in delivering quality output in various projects, leading to improved client satisfaction. - Successfully implemented several machine learning algorithms to develop predictive models for identifying customer churn, improving the accuracy of the model by 15% and contributing to a reduction in customer churn for the client. - Conducted exploratory data analysis and utilized data visualization techniques to identify trends and patterns in customer feedback, resulting in actionable insights for the client to improve their product and services. Utilized natural language processing techniques to build a sentiment analysis system, which provided valuable insights to the client for improving customer experience. - Contributed to the development of an image classification model using computer vision techniques, which helped automate the classification of images in a client's database, reducing manual labor and improving efficiency.
Samsung Electronics
Oct 20206 mosMar 2021

As a Samsung Prism Researcher at the Samsung R&D Bangalore Voice Intelligence group, I was responsible for developing anti-spoofing, speaker verification, and false wakeup solutions for IoT devices. This experience allowed me to gain a deep understanding of voice technology and contribute to the development of secure and user-friendly IoT devices.
GitHub
@vnykzhub

Vinayak’s AI Stack

antigravity
antigravity
claude code
claude code
claude
claude
codex
codex
cursor
cursor
deepseek
deepseek

Education

Deakin University

Master of Data Science (Global)

Deakin University

20232 yrs2025
Texas McCombs School of Business

Post Graduate Program, Artificial Intelligence and Machine Learning

Texas McCombs School of Business

20231 yr2024

Join Vinayak and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Vinayak Mathur on Tal

Headline

AI Engineer | Expert in Generative AI & Cloud Solutions | Nvidia Certified Generative AI Associate | Deloitte | ex-Cognizant | 4+ years of Experience

About

I’m an AI Software Engineer with 3 years of experience specializing in Generative AI, multi-agent frameworks, and MLOps. Over the course of my career, I’ve successfully: • Built multi-agent autonomous systems using Nvidia NIM that reduced operational overhead and achieved near-zero human intervention. • Improved audit workflows by 65% and boosted accuracy by 30% through advanced NLP and RAG (Retrieval Augmented Generation) pipelines. • Deployed cloud-native microservices with Kubernetes and Docker, hitting 99.9% uptime while cutting infrastructure costs by 40%. My passion lies in solving real-world problems by creatively blending cutting-edge AI with scalable cloud solutions. I love diving into complex challenges—whether that’s architecting enterprise tools for hands-free automation or optimizing deployments for maximum efficiency. Beyond coding, I focus on creating strategies that truly align with business objectives, delivering measurable impact and streamlined workflows. If you’re looking for someone who combines deep technical know-how with a solutions-driven mindset, I’m here to collaborate and drive your AI initiatives to new heights. Let’s connect and explore how we can innovate together!

Location

Hyderabad

AI stack

antigravity

claude code

claude

codex

cursor

deepseek

Work experience

AI and Data Science Engineer 1 at Deloitte, Nov 2025 - Present · 1 yr 1 mo, Hyderabad

Associate AI Engineer at Cognizant, Jul 2025 - Oct 2025 · 4 mos, Gurugram, Haryana, India, - Led the architecture and end-to-end development of a generative AI-powered data analyst for a global beverage leader. The agent automates BI workflows by translating user questions into SQL queries, data visualizations, and full analytical reports. - Engineered the system's core intelligence through a sophisticated prompt library, defining complex business rules for YTD analysis and market share calculations. - Built a resilient, multi-agent system (using LangGraph) where a central orchestrator delegates tasks to specialized agents for SQL generation, data visualization, and advanced Pandas-based calculations. - Integrated key Azure services, including Azure OpenAI for reasoning, Azure Blob Storage for stateful conversation persistence, and Azure SQL for live data querying.

Jr Data Scientist at Cognizant, Jul 2022 - Jul 2025 · 3 yrs 1 mo, Gurugram, Haryana, India, Multi-Agent Autonomous System Framework (Nvidia NIM, RAG, Python) • Architected and implemented a fully autonomous multi-agent system with Nvidia NIM, orchestrating specialized agents across diverse enterprise tools. • Developed a Linux system control agent for complex command execution without human intervention, and an enterprise communication agent handling Outlook/Teams interactions via Microsoft Graph API. • Engineered a development environment agent for hardware/software configuration, integrated a RAG pipeline for context-aware debugging, and set up agent-to-agent communication protocols for collaborative problem-solving. Automated Interview Platform (Generative AI, Kubernetes, Azure, AWS, GCP) • Created a full-stack automated interview solution leveraging LLMs (OpenAI GPT-4, Anthropic Claude), incorporating real-time speech processing. • Deployed Speech-to-Text and Text-to-Speech services on Kubernetes, achieving 99.9% uptime while reducing costs by 40%. • Implemented a multi-cloud strategy (Azure, AWS, GCP) with a unified CI/CD pipeline to streamline updates and resource allocation. Big 4 Financial Audit Automation & Analysis System (Generative AI, Python, Azure) • Developed a RAG-powered audit workflow for a Big 4 firm, automating lead sheet creation and analysis with GPT-3 and GPT-4. • Employed Python libraries (Openpyxl, Camelot, pdfPlumber) to parse financial data from PDFs, maintaining table integrity. • Built a user-friendly Streamlit interface that cut audit prep time by 65% and increased accuracy by 30%. Advanced Research Projects • Led comparative analysis of emerging LLMs (GPT, Claude, Llama, Gemini, Deepseek) to assess enterprise application viability. • Explored knowledge graph techniques for semantic relationship extraction from unstructured data. • Investigated prompt engineering, data generation, and system enhancements to optimize AI performance and reliability.

AI/ML Intern at Cognizant, Jan 2022 - Jul 2022 · 7 mos, - Specialized in AI/ML and worked on multiple projects involving AIML, NLP, and data analytics during GENC Next Internship at Cognizant. - Expanded skills in Pandas, Time Series Analysis, Statistical Data Analysis, AWS SageMaker, Amazon EC2, Apache Spark, Microsoft Azure, SQL, and Machine Learning through hands-on experience and training sessions. - Developed a strong foundation in data analysis and AI technologies, which helped in delivering quality output in various projects, leading to improved client satisfaction. - Successfully implemented several machine learning algorithms to develop predictive models for identifying customer churn, improving the accuracy of the model by 15% and contributing to a reduction in customer churn for the client. - Conducted exploratory data analysis and utilized data visualization techniques to identify trends and patterns in customer feedback, resulting in actionable insights for the client to improve their product and services. Utilized natural language processing techniques to build a sentiment analysis system, which provided valuable insights to the client for improving customer experience. - Contributed to the development of an image classification model using computer vision techniques, which helped automate the classification of images in a client's database, reducing manual labor and improving efficiency.

Samsung PRISM Researcher at Samsung Electronics, Oct 2020 - Mar 2021 · 6 mos, As a Samsung Prism Researcher at the Samsung R&D Bangalore Voice Intelligence group, I was responsible for developing anti-spoofing, speaker verification, and false wakeup solutions for IoT devices. This experience allowed me to gain a deep understanding of voice technology and contribute to the development of secure and user-friendly IoT devices.

Education

Master of Data Science (Global), Deakin University

Post Graduate Program, Artificial Intelligence and Machine Learning, Texas McCombs School of Business

Public links

github: https://github.com/vnykzhub