Vaibhav Patel, Developer in Dubai, United Arab Emirates
Vaibhav is available for hire
Hire Vaibhav

Vaibhav Patel

Verified Expert  in Engineering

Bio

Vaibhav is a skilled back-end engineer who specializes in building and fine-tuning large language models (LLMs) with Python. He delivers scalable AI solutions by leveraging advanced frameworks like LangChain and retrieval-augmented generation (RAG). With a strong foundation in LLM integration and optimization, he creates robust AI systems designed for high-performance applications.

Portfolio

ReturnQueen
Python, Large Language Models (LLMs), Natural Language Processing (NLP)...
Raxter
Artificial Intelligence (AI), Document Processing, Custom BERT, MySQL...
Lyearn
JavaScript, Data Architecture, API Architecture, REST APIs, Microservices...

Experience

  • Node.js - 6 years
  • Python - 6 years
  • Deep Learning - 5 years
  • Natural Language Processing (NLP) - 4 years
  • Retrieval-augmented Generation (RAG) - 2 years
  • Large Language Models (LLMs) - 2 years
  • C++ - 2 years
  • OpenAI GPT-4 API - 1 year

Availability

Part-time

Preferred Environment

Amazon Web Services (AWS), Python, Node.js, Large Language Models (LLMs), Retrieval-augmented Generation (RAG), C++

The most amazing...

...thing I've created an AI GitHub review system with RAG, performing RCA on code issues, suggesting fixes, and automating seamless integration into workflows.

Work Experience

Back-end Engineer

2021 - 2024
ReturnQueen
  • Designed and implemented a scalable back-end architecture using Python and Amazon EC2, enabling seamless horizontal scaling of microservices and improving system reliability.
  • Developed and optimized large-scale data pipelines for real-time data processing and analytics using Python, Apache Spark, and Kafka, enhancing data throughput and processing efficiency.
  • Architected a distributed system to handle single-digit millions of API requests per day using Kubernetes and Docker for container orchestration, improving load balancing and reducing response times.
  • Integrated Python-based ETL workflows with cloud storage solutions like Amazon S3 and Redshift to process and store large datasets, significantly reducing data processing times.
  • Deployed and managed fault-tolerant microservices using AWS Lambda and Amazon EC2 Auto Scaling, ensuring system reliability during high-traffic periods and maintaining high availability.
  • Built and optimized back-end APIs for high concurrency and large datasets, leveraging Python Flask and FastAPI, resulting in reduced latency and improved API response times under heavy load.
Technologies: Python, Large Language Models (LLMs), Natural Language Processing (NLP), OpenAI GPT-4 API, Retrieval-augmented Generation (RAG), Prompt Engineering, Generative Artificial Intelligence (GenAI), Pinecone, Vector Search, OpenAI

Research Engineer

2020 - 2021
Raxter
  • Created and deployed a research paper summarizer that takes the full text of the paper and classifies each sentence into various buckets, such as research goal, novelty, and limitations. Deployed the app on EC2 and AWS Lambda.
  • Contributed to the PDF parsing pipeline. Gathered multiple available PDF parsers and combined those outputs. Deployed on AWS.
  • Deployed figure extraction library on AWS EC2 with health check alarms, auto-scaling, and reporting.
  • Developed and deployed advanced NLP and text-to-speech models, enhancing the platform's capability to convert PDFs and web pages into audio notes with high accuracy and naturalness.
  • Optimized speech synthesis models using PyTorch, improving audio quality output by 30%, focusing on clarity and natural tone for multilingual support, including English, Spanish, and Mandarin.
  • Integrated AI models into the existing system architecture, enabling seamless functionality and user experience, which resulted in a 25% increase in user engagement.
  • Led rigorous testing and validation of NLP and text-to-speech models, achieving a 20% reduction in errors and mismatches, ensuring optimal performance and reliability.
  • Pioneered the use of serverless GPUs for model training and inference, reducing operational costs by 40% while accelerating model development cycles by 50%.
  • Implemented Docker containers for deploying AI models, enhancing the scalability and portability of the application across different environments.
Technologies: Artificial Intelligence (AI), Document Processing, Custom BERT, MySQL, SQLAlchemy, Recommendation Systems, Amazon Web Services (AWS), Python, Flask, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), SpaCy, Docker, Google Cloud Platform (GCP), Time Series, Time Series Analysis, Neural Networks, Data Analysis, Selenium, Data Engineering, Databricks, modal, Phonemes

Full-stack Developer

2018 - 2020
Lyearn
  • Created an internal npm library for reporting. Fetched and formatted data from Elastic Search, S3, and DynamoDB.
  • Worked on a new logic in Express.js, specifically Node.js, when an asset is granted or revoked to sub-account, which reduced the computation and database cost 100 fold.
  • Worked on the dark theme, live training, reporting, and live class on the platform.
Technologies: JavaScript, Data Architecture, API Architecture, REST APIs, Microservices, JavaScript 6, React, AWS Lambda, Amazon DynamoDB, Node.js, Azure, Business Requirements, Data Pipelines, SQL, Data Engineering

Research Engineer and Full-stack Engineer

2018 - 2019
Visionion
  • Provided services for prototyping, development, and deployment of machine learning algorithms.
  • Worked on image classification, object detection, and image super-resolution.
  • Contributed to an IoT project in Python and React, which uses serial communication and picoscope hardware.
Technologies: Object Detection, Optical Character Recognition (OCR), Amazon Web Services (AWS), Keras, PyTorch, TensorFlow, Machine Learning, Deep Learning, Python, Object Tracking, Video Analysis, Artificial Intelligence (AI), Data Pipelines, Data Analysis, Data Engineering

Machine Learning Engineer

2018 - 2018
Infocusp
  • Explored various deep learning algorithms to predict next frames using the existing user-drawn sketch.
  • Proposed a novel algorithm pipeline using optical flow, RNN, and CNN to solve the problem.
  • Implemented the algorithm using TensorFlow, Keras, and PyTorch. Trained and deployed the models on AWS EC2.
Technologies: Python, Computer Vision, Deep Learning, Artificial Intelligence (AI), NVIDIA CUDA, OpenCL, OpenCL/GPU, Data Analysis

Experience

GitHub Pull Requests Analysis System with RCA and RAG Integration

I developed an advanced GitHub-based solution that integrates large language models (LLMs) and retrieval-augmented generation (RAG) using LangChain to analyze and review pull requests (PRs). The system performs automated root cause analysis (RCA) on issues linked to PRs, leveraging multi-agent systems for comprehensive analysis and resolution suggestions. It also has direct access to repositories, enabling it to suggest and implement fixes autonomously, streamlining the development process for complex systems.

Classification of Plant Disease Using Convolutional Neural Networks

As part of the Summer Innovation Challenge 2017 hosted by the Gujarat State Government, I trained a CNN with transfer learning to classify a plant's disease from its images. I implemented flexible, fast code that supports training on the line.

Motion-based Image Super Resolution

Proposed and implemented a general-purpose deep learning architecture that can learn multi-image to multi-image mapping. I also investigated its application in image registration, image super-resolution, and photometric stereo.

Tone Mapping HDR Images

Tone-mapped high dynamic range (HDR) images using the generative adversarial network (GAN). I proposed a GAN architecture to tone-map HDR images from the output of multiple TMO algorithms present. I used an extension of an image translation network called pix2pix.

Rice Classification Using CNNs

Proposed a convolutional neural network architecture with transfer learning that outperforms state-of-the-art methods. I also trained a 5-class model for classifying basmati rice using 4,000 training images.

Education

2014 - 2018

Bachelor's Degree in Computer Science

DA-IICT - Gandhinagar, India

Certifications

AUGUST 2017 - PRESENT

Neural Networks for Machine Learning

Coursera

SEPTEMBER 2016 - PRESENT

Machine Learning

Coursera

Skills

Libraries/APIs

Node.js, PyTorch, React, SpaCy, TensorFlow, Keras, Scikit-learn, NumPy, SciPy, OpenCV, Matplotlib, SQLAlchemy, LSTM, Pandas, MobX, REST APIs

Tools

Jupyter, Git, Amazon Elastic Container Service (ECS), Apache Airflow

Languages

Python, SQL, Python 3, JavaScript 6, JavaScript, CSS, HTML, HTML5, C++, TypeScript

Platforms

Amazon Web Services (AWS), AWS Lambda, NVIDIA CUDA, Docker, Databricks, Azure, Google Cloud Platform (GCP)

Frameworks

Flask, Selenium, Caffe, Redux, Angular, OpenCL

Paradigms

Microservices Architecture, Serverless Architecture, Microservices, API Architecture

Storage

Amazon DynamoDB, Elasticsearch, MySQL, Data Pipelines, PostgreSQL, PostGIS

Other

Deep Learning, Computer Vision, Machine Learning, Natural Language Processing (NLP), Artificial Intelligence (AI), Generative Pre-trained Transformers (GPT), Data Analysis, Data Engineering, BERT, Large Language Models (LLMs), Text to Speech (TTS), OpenAI GPT-4 API, OpenAI GPT-3 API, Retrieval-augmented Generation (RAG), LangChain, OpenAI, Deep Neural Networks (DNNs), Convolutional Neural Networks (CNNs), Pattern Recognition, Image Classification, Data Science, Computer Vision Algorithms, Recommendation Systems, Prompt Engineering, Pinecone, Vector Search, Mathematics, Generative Adversarial Networks (GANs), Recurrent Neural Networks (RNNs), Big Data, Data Architecture, Custom BERT, Document Processing, Optical Character Recognition (OCR), Object Detection, Unsupervised Learning, Object Tracking, Video Analysis, OpenCL/GPU, Time Series, Time Series Analysis, Neural Networks, Business Requirements, modal, Phonemes, GPU Computing, Serverless GPUs, Image Processing, Multi-agent Systems, Generative Artificial Intelligence (GenAI)

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