Vishnu is currently unavailable

Vishnu Anurag Thonukunoori

Verified Expert in Engineering

Artificial Intelligence and Machine Learning Engineer and Developer

Franklin Township, NJ, United States

Toptal member since August 8, 2025

Expertise

Machine Learning Neural Network Artificial Intelligence Serverless Data Analysis Deep Learning Prompt Engineering LLM RAG OpenAI Python TensorFlow AWS Lambda React.js

Bio

Vishnu is an AI/ML engineer with 5+ years of experience building scalable machine learning and generative AI solutions for enterprises. He specializes in LLMs, RAG pipelines, and AWS-based deployments and has had proven success in NLP, prompt engineering, and vector search. Vishnu is trusted by clients across insurance, eCommerce, and sustainability sectors for delivering production-grade AI systems that drive business value.

Portfolio

Aapoon

AWS Bedrock AgentCore, Amazon SageMaker, Claude, Pinecone, AWS Glue, AWS Lambda...

The Hanover Insurance Group

BERT, Amazon SageMaker, Elasticsearch, AWS AppSync, Open-source LLMs...

1800Spirits

TensorFlow, XGBoost, Logistic Regression, AI Model Training, Amazon EC2...

Experience

Machine Learning - 5 years
Python 3 - 5 years
TensorFlow - 5 years
AI Model Training - 5 years
Model Deployment - 5 years
Amazon SageMaker - 5 years
AWS Bedrock AgentCore - 2 years
LangChain - 2 years

Preferred Environment

Python 3, TensorFlow, Amazon SageMaker, AWS Bedrock AgentCore, LangChain, Hugging Face Transformers, Agentic AI, Generative Artificial Intelligence (GenAI), Machine Learning, Artificial Intelligence (AI)

The most amazing...

...thing I've built is a RAG-based GenAI assistant using LangChain, OpenAI, and Pinecone. It cut response time by 30% and automated NLU workflows.

Work Experience

Machine Learning Engineer | Full-stack Developer

2024 - 2025

Aapoon

Built a RAG-based assistant using LangChain and OpenAI, reducing response time by 30%.
Integrated FAISS and Pinecone for hybrid retrieval in chatbot workflows.
Built reusable React components for group sync, polling, and messaging interfaces, improving front-end load performance by 25%.
Engineered back-end APIs in Python using AWS Lambda for policy management, claim status, and agent workflows.
Created interactive dashboards using Streamlit to visualize usage metrics and user feedback.
Containerized services using Docker and deployed to AWS Lambda via CI/CD pipelines.
Developed and maintained dynamic web applications using PHP, integrating with MySQL databases and RESTful APIs.
Optimized back-end logic, improved application performance, and implemented secure authentication and session management.

Technologies: AWS Bedrock AgentCore, Amazon SageMaker, Claude, Pinecone, AWS Glue, AWS Lambda, React, Amazon S3 (AWS S3), MongoDB, CI/CD Pipelines, Docker, Machine Learning, Python, Natural Language Processing (NLP), Artificial Intelligence (AI), Large Language Models (LLMs), Performance, Agentic AI, Generative Artificial Intelligence (GenAI), Data Analytics, Data Science, Data Analysis, AI Agents, APIs, Machine Learning Operations (MLOps), Back-end, Object-oriented Programming (OOP), Amazon Web Services (AWS), Retrieval-augmented Generation (RAG), Package Distribution, Statistics, AI Chatbots, Document Processing, OpenAI, Chatbots, Object Detection, Recommendation Systems, Vector Databases, Leadership, Chatbot Conversation Design, Conversational AI, Conversational Agent, Model Tuning, Serverless, Node.js, Document Parsing, Full-stack, Front-end Development, JavaScript, Cloud Infrastructure, Cursor AI, Microsoft Copilot Studio, Microsoft Copilot, Generative Pre-trained Transformers (GPT), Meta Llama, PostgreSQL, Automation, Data Scraping, Scraping, Big Data, Fine-tuning, AI Voice Agents, Architecture

Machine Learning Engineer | Full-stack Developer

2023 - 2024

The Hanover Insurance Group

Engineered a clause classification pipeline using BERT fine-tuned on insurance policy data.
Built back-end sync services in Node.js and Python to integrate and normalize inventory from 6 POS systems into PostgreSQL.
Developed schema-first GraphQL endpoints using AWS AppSync to reduce client-side logic and replace legacy REST APIs.
Built and deployed ML workflows on AWS SageMaker and GraphQL APIs via AppSync.
Executed model evaluation using custom scoring metrics and AWS Sagemaker tools and services.
Worked with actuarial data to support model features related to claims history, loss ratios, and premium trends.

Technologies: BERT, Amazon SageMaker, Elasticsearch, AWS AppSync, Open-source LLMs, Model Evaluation, Machine Learning, Python, Natural Language Processing (NLP), Artificial Intelligence (AI), Large Language Models (LLMs), Performance, Agentic AI, Generative Artificial Intelligence (GenAI), Data Analytics, Data Science, Data Analysis, AI Agents, APIs, Machine Learning Operations (MLOps), Back-end, Object-oriented Programming (OOP), Amazon Web Services (AWS), Retrieval-augmented Generation (RAG), Package Distribution, Statistics, AI Chatbots, Document Processing, OpenAI, Chatbots, Recommendation Systems, Vector Databases, Leadership, Chatbot Conversation Design, Conversational AI, Conversational Agent, Model Tuning, Serverless, Node.js, Document Parsing, Full-stack, Predictive Modeling, Front-end Development, JavaScript, Cloud Infrastructure, Cursor AI, Microsoft Copilot Studio, Microsoft Copilot, Generative Pre-trained Transformers (GPT), Meta Llama, PostgreSQL, Automation, Data Scraping, Scraping, Big Data, Fine-tuning, AI Voice Agents, Architecture, Pricing, Pricing Models

Machine Learning Engineer

2021 - 2023

1800Spirits

Built a collaborative filtering engine for personalized product recommendations.
Trained a neural net that boosted top-5 recommendation accuracy by 22%.
Processed 1+ million SKU events using batch ETL workflows and Amazon S3 based feature stores.
Created churn prediction models using XGBoost and logistic regression.
Delivered real-time inference using AWS Lambda and API Gateway.

Technologies: TensorFlow, XGBoost, Logistic Regression, AI Model Training, Amazon EC2, Amazon S3 (AWS S3), AWS Lambda, Machine Learning, Python, Artificial Intelligence (AI), Large Language Models (LLMs), Performance, Data Analytics, Data Science, Data Analysis, Supervised Machine Learning, Labeling, Data Labeling, APIs, Machine Learning Operations (MLOps), Image Processing, Back-end, Object-oriented Programming (OOP), Amazon Web Services (AWS), Package Distribution, Statistics, Document Processing, Recommendation Systems, Leadership, Model Tuning, Serverless, Node.js, Document Parsing, XML, SQL, Web Scraping, Full-stack, Predictive Analytics, Predictive Modeling, Time Series Forecasting, API Integration, Front-end Development, JavaScript, Cloud Infrastructure, PostgreSQL, Data Scraping, Scraping, Big Data, Architecture, Pricing, Pricing Models

Machine Learning Engineer

2020 - 2021

Waste Management

Built LSTM models that predicted service delays from vehicle telemetry data.
Reduced SLA violations by 18% by delivering predictive insights to dispatch teams.
Built ETL pipelines to process GPS data from trucks using AWS Glue.
Clustered driver behavior using telemetry patterns to optimize route planning.
Created performance dashboards using Athena and QuickSight for SLA monitoring.

Technologies: PyTorch, TensorFlow, AI Model Training, Amazon EC2, Amazon S3 (AWS S3), LSTM Networks, Architecture best practices, AWS Glue, Amazon DynamoDB, Clustering, Machine Learning, Python, Artificial Intelligence (AI), Large Language Models (LLMs), Performance, Data Analytics, Data Science, Data Analysis, Supervised Machine Learning, Labeling, Data Labeling, APIs, Machine Learning Operations (MLOps), Image Processing, Back-end, Object-oriented Programming (OOP), Amazon Web Services (AWS), Package Distribution, Statistics, Model Tuning, Serverless, Node.js, Document Parsing, XML, SQL, Full-stack, Predictive Analytics, Predictive Modeling, Time Series Forecasting, API Integration, Front-end Development, JavaScript, Cloud Infrastructure, PostgreSQL, Architecture

Experience

LLM-powered Chatbot to Understand User Mental Health and Well-being

https://dagshub.com/Omdena/HyderabadIndiaChapter_MentalHealthWellbeingFomoSocialMedia/src/team-consolidated/src

An advanced LLM-powered chatbot aimed at analyzing social media activity to detect early signs of mental health issues like anxiety, depression, FOMO, and suicidal ideation. Using Hugging Face Transformers and sentiment classification models, the chatbot was fine-tuned on Reddit data and deployed with modular Python scripts.

I implemented mental health-related intent recognition, integrated risk-level scoring logic, and helped visualize user well-being trends. Based on the identified patterns, the chatbot offered contextual recommendations and resource links.

ML Workflow for Scones Unlimited on Amazon SageMaker

https://github.com/Vishnu2001-tech/Udacity_Project_top_500

Developed an end-to-end ML workflow to automate vehicle-based delivery routing for Scones Unlimited. The project's core involved training an image classification model using deep learning techniques on Amazon SageMaker to identify the type of delivery vehicle, i.e., bike, motorcycle, or van, from input images. This helped assign deliveries based on proximity and vehicle efficiency.

Landmark Classification and Tagging for Social Media

https://github.com/Vishnu2001-tech/Udacity_Project_Landmark

Built a deep learning pipeline to automatically classify and tag images based on well-known landmarks using convolutional neural networks (CNNs). The model was trained using transfer learning on a pre-trained CNN architecture, fine-tuned on a custom landmark image dataset. This enabled accurate prediction of landmark locations and tagging for use in social media applications. I also developed a Streamlit-based front end to display results and confidence scores for end-user testing.

Education

2019 - 2023

Bachelor's Degree in Mechanical Engineering

Jawaharlal Nehru Technological University, College of Engineering Hyderabad - Hyderabad, Telangana, India

Certifications

JANUARY 2024 - JANUARY 2027

TensorFlow Developer Certificate

TensorFlow

JUNE 2023 - JUNE 2026

AWS Certified Solutions Architect - Associate

Amazon Web Services Training and Certification

Skills

Libraries/APIs

TensorFlow, React, PyTorch, Hugging Face Transformers, Node.js, Scikit-learn, Keras, XGBoost

Tools

Amazon SageMaker, Git, Microsoft Copilot, Jupyter, PyCharm, Claude, AWS Glue, AWS AppSync, AWS IAM, AWS Step Functions

Languages

Python 3, Python, SQL, XML, JavaScript

Platforms

AWS Lambda, Amazon Web Services (AWS), Docker, Microsoft Copilot Studio, Visual Studio Code (VS Code), Amazon EC2

Storage

PostgreSQL, Elasticsearch, Amazon S3 (AWS S3), MongoDB, Amazon DynamoDB

Paradigms

Object-oriented Programming (OOP), Automation

Frameworks

Streamlit

Other

Machine Learning, Deep Learning, LangChain, Neural Networks, Prompt Engineering, Artificial Intelligence (AI), Large Language Models (LLMs), Agentic AI, Generative Artificial Intelligence (GenAI), Data Science, Data Analysis, Supervised Machine Learning, Labeling, Data Labeling, AI Agents, Machine Learning Operations (MLOps), Retrieval-augmented Generation (RAG), Package Distribution, Statistics, OpenAI, Leadership, Serverless, Cloud Infrastructure, Generative Pre-trained Transformers (GPT), Fine-tuning, AWS Bedrock AgentCore, Computer Vision, AI Model Training, Model Deployment, CI/CD Pipelines, Natural Language Processing (NLP), Performance, Data Analytics, APIs, Image Processing, Back-end, AI Chatbots, Document Processing, Chatbots, Object Detection, Recommendation Systems, Vector Databases, Chatbot Conversation Design, Conversational AI, Conversational Agent, Model Tuning, Document Parsing, Web Scraping, Full-stack, Predictive Analytics, Predictive Modeling, Time Series Forecasting, API Integration, Front-end Development, Cursor AI, Meta Llama, Data Scraping, Scraping, Big Data, Architecture, Pricing, Pricing Models, CAD/CAM, Data Preprocessing, Model Evaluation, Pinecone, BERT, Open-source LLMs, AWS Cloud Architecture, Virtual Private Cloud (VPC), Cost Reduction & Optimization (Cost-down), Architecture best practices, Cloud Architecture, Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Image Classification, ML Deployment, Logistic Regression, LSTM Networks, Clustering, Transfer Learning, ETL Pipelines, AI Voice Agents

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.

Share your needs

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

Choose your talent

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

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