Rishab is available for hire

Rishab Pal

Verified Expert in Engineering

AI Architect and Staff Machine Learning Developer

Bengaluru, Karnataka, India

Toptal member since November 10, 2022

Expertise

Machine Learning Artificial Intelligence Integration NLP Data Science Cloud Engineering LLM Deep Learning Computer Vision Neural Network Code Review DeepSeek Voiceflow

Bio

Rishab architects AI systems that don't just work in theory; they ship to production and drive measurable results. With 8+ years of end-to-end ML engineering across LLMs, computer vision, NLP, and multimodal AI, he brings research depth and engineering discipline to every engagement. Rishab consistently delivers systems that are fast, accurate, and built to scale.

Portfolio

Aon - Global - Singapore

Artificial Intelligence (AI), Python, Retrieval-augmented Generation (RAG)...

Discover Healing, Inc.

ChatGPT, AI Integration, JavaScript, Large Language Models (LLMs), Fine-tuning...

Invisible Technologies

Prompt Engineering, Artificial Intelligence (AI)...

Experience

Python - 9 years
Data Science - 8 years
Natural Language Processing (NLP) - 7 years
Data Analytics - 6 years
Deep Learning - 5 years
Software Deployment - 5 years
Statistical Modeling - 5 years
DeepSeek - 2 years

Preferred Environment

Machine Learning, Deep Learning, Data Science, Natural Language Processing (NLP), Computer Vision, Python, Artificial Intelligence (AI), Generative Artificial Intelligence (GenAI), Machine Learning Operations (MLOps)

The most amazing...

...accomplishment I've had was showcasing my project at the CES—International Consumer Electronics Show, the technology sector's flagship trade fair.

Work Experience

GenAI Architect

2026 - PRESENT

Aon - Global - Singapore

Engineered a natural-language-to-SQL chatbot using LangGraph and OpenAI that decomposes complex user queries into dependency-ordered sub-query DAGs and executes independent branches in parallel against Databricks SQL warehouses.
Designed a hybrid SQL generation pipeline combining pre-validated semantic-similarity-matched templates as tools with LLM-generated dynamic SQL fallback, reducing hallucination risk for known query patterns to near zero.
Implemented a 5-layer accuracy safeguard system including SQL validation nodes, automated retry loops (up to three attempts), read-only SELECT enforcement, schema-aware JOIN resolution, and post-execution result sanity checks.
Built a dependency-aware query decomposition engine that breaks complex questions into a DAG of sub-queries, executing independent branches in parallel via LangGraph's Send API and injecting upstream results as context for dependent branches.

Technologies: Artificial Intelligence (AI), Python, Retrieval-augmented Generation (RAG), Generative Artificial Intelligence (GenAI), OpenAI, Large Language Models (LLMs), LangChain, Ollama, LlamaIndex, SQL, Databricks, Amazon Web Services (AWS), Azure, Docker, Kubernetes

Senior Full-stack AI Engineer

2025 - PRESENT

Discover Healing, Inc.

Built a RAG-powered AI chatbot (MedPal) for Discover Healing using OpenAI Agents SDK and Vector Store, enabling accurate Q&A over energy healing content (Emotion Code, Body Code, Belief Code).
Implemented multi-layer guardrails — regex and OpenAI Moderation API for input, plus LLM-based hallucination and FTC policy judges for output, ensuring no medical claims or healing promises reach users.
Integrated Zoho SalesIQ live chat via a FastAPI webhook with RSA signature verification, multi-step name/email intake flow, and session persistence for seamless live-chat handoff.
Deployed dual-surface architecture on Google Cloud Run — a Streamlit admin UI (with document management) and a production webhook server, both containerized via Docker and managed with uv.

Technologies: ChatGPT, AI Integration, JavaScript, Large Language Models (LLMs), Fine-tuning, Python, APIs, Zoho, AI Chatbots, AI Voice Agents, Conversational AI

Senior AI Developer and Prompt Engineer

2024 - 2026

Invisible Technologies

Developed hallucination detection and content verification systems using LangChain agents and LangGraph, creating ensemble techniques that reduced inaccuracies by 65% while maintaining strict data privacy protocols for enterprise clients.
Engineered automated prompt frameworks for claim extraction and verification using OpenAI Assistants and custom Python tooling, enabling automated quality assurance processes that scaled to process 30,000+ content pieces daily.
Implemented and fine-tuned Llama models as verification agents (LLM as a judge), developing novel techniques for cost-effective factuality assessment that maintained 92% accuracy while reducing API costs by 72%.
Built a novel automated red teaming system that uses two adversary LLM (a planner and an attacker) to attack a victim LLM. A classifier judges whether or not conversations contain harmful content.
Proposed Automatic Prompt Engineer (APE) for automatic instruction generation and selection, drawing inspiration from classical and human approaches to prompt engineering and demonstrating the significant impact of prompt quality on task performance.

Technologies: Prompt Engineering, Artificial Intelligence (AI), Natural Language Processing (NLP), Machine Learning, Data Science, Python, PyTorch, TensorFlow, Cloud

AI Architect and Machine Learning Engineer

2025 - 2025

Mario Peng Lee

Implemented a production-ready MCP system with LangChain and LangGraph, replacing a legacy forms-based chatbot while maintaining 100% functional parity via parallel execution strategy.
Built seven specialized MCP tools (SQL generation, forecasting, and weather) with sophisticated restaurant business logic for sales analytics, projections, and operational intelligence.
Deployed XGBoost and statistical forecasting models, achieving a 56% accuracy improvement, and integrated with the MCP agent for real-time sales predictions and impact analysis.
Developed a comprehensive event impact analyzer with service-type specific multipliers, location proximity effects, and 20+ event types affecting restaurant demand patterns.
Executed a zero-downtime "Parallel Evolution" approach with separate MCP processors, dedicated test suites, and deployment scripts enabling gradual production transition.

Technologies: Artificial Intelligence (AI), Model Context Protocol (MCP), Retrieval-augmented Generation (RAG), Database Optimization, Architecture, Chatbots, AI Chatbots, Model Context Protocol

AI Developer and Architect

2025 - 2025

Tomorrow People

Developed and optimized enterprise-grade prompt engineering frameworks using LangChain and LangGraph, resulting in a 40% improvement in content generation quality for sales and marketing materials across multiple industry verticals.
Led knowledge management initiatives including fine-tuning Llama models on domain-specific data, building structured knowledge graphs, and designing retrieval-augmented generation (RAG) systems that improved content relevance by 35%.
Architected and implemented custom AI agent systems with CrewAI that enhanced sales workflows, integrating with Claude and ChatGPT via APIs while maintaining comprehensive documentation and prompt libraries.
Collaborated with sales and marketing teams to analyze prompt performance metrics, conduct A/B testing on different prompt strategies, and iteratively refine AI solutions to meet specific business requirements.

Technologies: Prompt Engineering, AI Prompts, Artificial Intelligence (AI), APIs, API Integration, Claude, ChatGPT

CTO/LLM Expert

2024 - 2024

Haadi

Successfully delivered POC for Toptal client: AI-powered mental health assistant with 3D avatar interface and voice interactions mimicking real therapist sessions.
Engineered Multi-Agent RAG conversational system with neuroscientist expertise emulation and vector-based chat history for clinical coherence, like therapeutic structures such as CBT or DBT.
Implemented affective computing to detect user emotional states and dynamically retrieve contextually relevant information from domain knowledge bases.
Achieved seamless integration of evidence-based clinical responses with naturally empathetic interactions through advanced AI architecture.

Technologies: Artificial Intelligence (AI), Large Language Models (LLMs), ChatGPT, Deep Learning, Minimum Viable Product (MVP), Speech-to-Text (STT), LangChain, Multimodal Models

Lead AI Developer

2024 - 2024

Solstice Health Inc

Extracted key medical claims from research documents and materials published by U.S. pharmaceutical companies, meeting FDA requirements.
Used extracted data to create presentations for customers and clinicians, supporting sales pitches for new drugs.
Built the solution on LangChain, with fine-tuned models and prompt engineering.
Delivered over 90% successful claim extraction based on expert reviews and analysis.

Technologies: Python, Large Language Models (LLMs), Fine-tuning, ChatGPT, Mistral AI, Llama 2, Llama 3, Retrieval-augmented Generation (RAG), LangChain, AI Agents, DeepSeek, Claude, Gemini

Lead AI Developer

2024 - 2024

Flirt Technologies, Llc

Developed AI-powered social media app features for personalized content creation using Stable Diffusion and Deepgram.
Enabled near-real-time immersive experiences with professional model photos, videos, and voice cloning.
Used machine learning techniques like low-rank adaptation (LoRA) based on PEFT, model quantization, checkpoint merging, and other related approaches to achieve state-of-the-art (SOTA) optimization and performance.

Technologies: Artificial Intelligence (AI), Large Language Models (LLMs), Machine Learning, Stable Diffusion, OpenAI, Text to Image, Claude, Gemini

Lead Data Scientist

2023 - 2024

Makro

Optimized retail product pricing by employing an ML demand prediction model trained on time-series historical sales data. Conducted data analysis and stakeholder management, and designed an A/B test framework.
Obtained state-of-the-art outcomes in campaign banner creation through the use of Stable Diffusion by experimenting with various hyperparameter settings, performing fine-tuning, merging checkpoints, and conducting meticulous prompt engineering.
Developed a lifecycle segmentation model to classify customers into segments such as stable, declining, growing, and at-risk of churn based on transaction patterns, enabling targeted business strategies.
Automated the identification of target customer sets for marketing campaigns using the existing recommendation model, tripling participation rates and achieving a 20% win-back of churned customers.

Technologies: AI Agents, Python, A/B Testing, Large Language Models (LLMs), Databricks, Time Series Forecasting, Learning to Rank (LTR), Multi-armed Bandits (MABs), Pandas, OpenAI, SQL, Neo4j, Graph Databases, Vector Databases, Pinecone

AI/ML Expert

2023 - 2024

Nolea Technology Ltd

Built an MVP for the recommendation system to match healthcare companies with clinical specialists in mental healthcare, allied healthcare, orthopedics, etc.
Re-engineered the existing recommendation engine using SOTA NLP approaches and machine learning models like named-entity recognition (NER), LayoutLM, and LayoutParser, and built custom doc embeddings.
Designed a cloud-native trainable and scalable architecture, previously built on a rule engine that lacked robustness for their business model.

Technologies: Python, Scikit-learn, Pandas, Amazon Web Services (AWS), Artificial Intelligence (AI), Machine Learning, Recommendation Systems, Retrieval-augmented Generation (RAG), Generative Artificial Intelligence (GenAI), Azure AI Studio, ChatGPT, OpenAI Assistants API, OpenAI API, LangChain, Claude, Gemini

LLM Chatbot Developer

2023 - 2023

Revibe AI

Delivered Voiceflow-based mental health assistant POC with advanced user understanding, custom knowledge base suggestions, and human-like conversational capabilities.
Developed RAG-powered chatbot replicating neuroscientist behavior with empathetic user engagement and contextually appropriate clinical responses.
Integrated LangChain framework to enrich the vector store with proprietary neuroscience and psychological datasets for enhanced LLM response accuracy.
Architected an intelligent conversation system combining domain expertise simulation with personalized therapeutic interactions and knowledge retrieval.

Technologies: Generative Pre-trained Transformers (GPT), Python, Chatbot Conversation Design, API Integration, Natural Language Understanding (NLU), Natural Language Processing (NLP), Language Models, Artificial Intelligence (AI), Voiceflow, Cognitive Behavioral Therapy (CBT), Zapier, Machine Learning, Chatbots, Retrieval-augmented Generation (RAG), Generative Artificial Intelligence (GenAI), Azure AI Studio, LangChain, Integrated Development Environments (IDE), ChatGPT, OpenAI Assistants API, OpenAI API, AI Agents, Azure SQL Databases, Claude, Gemini

ML Expert | Data Scientist

2023 - 2023

Zachary Gidwitz

Built a recommendation engine for an addiction recovery app that incorporates user demographics, mental state, and other metadata from app interaction to suggest quotes and prayers relevant to the user's state of mind.
Fine-tuned the OpenAI GPT3 model and used Langchain to build a sequence of prompts where each preceding answer is included in the subsequent question. I added memory and linked the vector database quickly.
Provided several offline and online metrics to track the live performance of the model and the recommendation system.

Technologies: Data Science, Machine Learning, Python, Recommendation Systems, Data Strategy, Data, Chatbots, OpenAI GPT-3 API, Jupyter, Cloud, OpenAI GPT-4 API, Chatbot Conversation Design, Integration, Programming, AI Programming, Data Cleaning, Unstructured Data Analysis, Airtable, Machine Learning Automation, Project Management, Statistical Analysis, Data Reporting, Data Cleansing, Prompt Engineering, Retrieval-augmented Generation (RAG), Generative Artificial Intelligence (GenAI), Azure AI Studio, LangChain, ChatGPT, OpenAI Assistants API, OpenAI API, AI Agents, Azure SQL Databases

Machine Learning Engineer

2023 - 2023

Odem Global Pty Ltd

Trained causal language model (CLM) with DeepSpeed on Bittensor, a peer-to-peer open-source protocol powering a globally distributed decentralized neural network.
Used a mountain dataset (800+ GB) to train generative LLMs using DeepSpeed, which has lower network validation loss and thereby increases the model's reward, benefiting the client.
Contributed to hyper-parameter tuning and revamped the default PyTorch data loader for faster dataset loading. Also, validated the models' performance on the validation set to ensure it meets the desired performance metrics.
Fine-tuned models like GPT-J-Neo, Llama, MPT, Vicuna, Koala, Cerebras, to mention a few.

Technologies: Language Models, Machine Learning, Fine-tuning, Causal Inference, DeepSpeed, OpenAI GPT-3 API, Jupyter, Cloud, Integration, Programming, AI Programming, Data Cleaning, Unstructured Data Analysis, Machine Learning Automation, Large Language Models (LLMs), Project Management, Regex, Statistical Analysis, Data Reporting, Data Cleansing, Prompt Engineering, Chatbots, Retrieval-augmented Generation (RAG), Azure SQL Databases

AI Expert

2023 - 2023

Rich Lemon Apps FZE LLC

Developed a user sticker model using Stable Diffusion and ControlNet for a given style.
Experimented with different configurations of parameters to understand the effect on the results and achieved state-of-the-art results.
Explored different methods like checkpoint merging, trained a stickers model, and then fine-tuned that model on Portrait Face, and combined sticker style and user image training.

Technologies: Artificial Intelligence (AI), Deep Learning, Image Processing, Stable Diffusion, Computer Vision, ControlNet, 3D, Microservices Architecture, Data Scientist, Jupyter, Cloud, Integration, Programming, AI Programming, Data Cleaning, Machine Learning Automation, Image Analysis, Project Management, Statistical Analysis, Data Reporting, Data Cleansing, Prompt Engineering, Image Generation, Text to Image, Multimodal GenAI, OpenAI API, Azure SQL Databases

Senior ML Expert | Data Scientist

2021 - 2023

EagleView

Revamped roof measurement automation for US houses by engineering a high-performance Line Detection model on aerial imagery to detect roof, wall, and ground lines from 28% to 59%.
Developed segmentation models for roof and wall facet detection while handling obstruction, achieving 86% accuracy.
Improvised complex PyTorch models to facilitate ONNX and TensorRT export for efficient deployment using NVIDIA Triton Server.
Orchestrated microservices using Step Functions, AWS Lambda, and Amazon SageMaker endpoints, improving response time by 25%. This was followed by simplifying maintenance and support, focusing on maximizing code reuse.

Technologies: Python, Machine Learning Operations (MLOps), PyTorch, Python API, Docker, Computer Vision, GPU Computing, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), Data Science, SQL, Neural Networks, Data Visualization, Unsupervised Learning, Code Review, Remote Team Leadership, Graphics Processing Unit (GPU), Transfer Learning, Transformers, Sequence Models, Image Retrieval, Image Recognition, Classification, Entity Extraction, Siamese Neural Networks, Artificial Neural Networks (ANN), Convolutional Neural Networks (CNNs), TensorBoard, Generative Adversarial Networks (GANs), Linux, Language Models, Python 3, Recommendation Systems, Algorithms, XGBoost, Data Extraction, Data Manipulation, Snowflake, Analytics, MySQL, Amazon Web Services (AWS), Data Analysis, Version Control Systems, Communication, Data Engineering, Regression, RAPIDS, APIs, Amazon SageMaker, NoSQL, Image Search, AI Design, Jupyter Notebook, Flask, NVIDIA TensorRT, Microservices Architecture, Data Scientist, Jupyter, Cloud, Anomaly Detection, Integration, Programming, AI Programming, Bitbucket, Data Cleaning, Unstructured Data Analysis, Machine Learning Automation, Kubernetes, Amazon Elastic Container Service (ECS), Image Analysis, Large Language Models (LLMs), Project Management, Statistical Analysis, Data Reporting, Data Cleansing, You Only Look Once (YOLO), Databricks, Azure SQL Databases

Senior Software Engineer/Machine Learning (ML)

2020 - 2021

Airtel India

Designed and built a chatbot, leveraging HuggingFace transformers with DeepSpeed to resolve queries for over 400 million of Airtel's monthly active users. Deployed successfully, handling over 10 million daily conversations with 85% accuracy.
Improved the accuracy of face recognition, matching, and liveness-check models. Added face segmentation, blur and forgery detection, real-time blacklisting, and face search. Enhanced service efficiency and reduced manual effort by over 25%.
Inculcated best MLOps practices on the team to lower technical debt and adapt workflow that requires little to no manual intervention for simulations, testing, deployment, and monitoring.

Technologies: Data Analytics, Statistical Modeling, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), Python 3, Docker, Bash Script, Data Science, SQL, Optical Character Recognition (OCR), Neural Networks, Optimization, Clustering, Code Review, Remote Team Leadership, Graphics Processing Unit (GPU), Transfer Learning, Transformers, Sequence Models, Image Retrieval, Image Recognition, Classification, Entity Extraction, Recurrent Neural Networks (RNNs), Artificial Neural Networks (ANN), Convolutional Neural Networks (CNNs), TensorBoard, Generative Adversarial Networks (GANs), Topic Modeling, Linux, Language Models, PostgreSQL, Recommendation Systems, Algorithms, Generative Pre-trained Transformer 3 (GPT-3), XGBoost, Data Extraction, Data Manipulation, Snowflake, Analytics, MySQL, Amazon Web Services (AWS), Handwriting Recognition, Data Analysis, Version Control Systems, Communication, Data Engineering, Regression, RAPIDS, APIs, Amazon SageMaker, NoSQL, Image Search, Amazon Rekognition, AI Design, Jupyter Notebook, Flask, NVIDIA TensorRT, Microservices Architecture, Data Scientist, Chatbots, Jupyter, Cloud, Anomaly Detection, OpenAI GPT-4 API, Chatbot Conversation Design, Integration, Programming, AI Programming, Bitbucket, Data Cleaning, Unstructured Data Analysis, Machine Learning Automation, Kubernetes, Amazon Elastic Container Service (ECS), Image Analysis, Large Language Models (LLMs), Project Management, Statistical Analysis, Data Reporting, Data Cleansing, Big Data, Databricks, Azure SQL Databases

ML Engineer

2019 - 2020

Trantor Software

Designed and deployed an OCR solution that outperformed the accuracy of Google OCR by 12% as benchmarked on the ICDAR 2013 dataset. Reduced the overall cost from an estimated $1 million to less than $50,000 per year.
Prototyped a natural language processing (NLP)--based bot using recurrent neural networks to parse comments on the website, assess sentiments, and classify them into themes. Enabled standard responses and activation of the needed systems to respond.
Worked on a trained convolutional neural networks model using INT8 fixed-point arithmetic. Designed algorithms to decide the Q. Collaborated to enable classifiers, detectors, generative adversarial networks (GANs), and segmentation networks.
Trained a software categorization and brand recognition tool using word2vec, N-grams embeddings, and kNN classifier. This tool was capable of matching product names and titles with descriptions.
Managed quantization-aware training of convolutional neural networks and worked on fine-tuning and post-training quantization.

Technologies: Machine Learning, Deep Learning, Data Science, SQL, Artificial Intelligence (AI), Optical Character Recognition (OCR), Data Visualization, Optimization, Unsupervised Learning, Clustering, Code Review, Remote Team Leadership, Graphics Processing Unit (GPU), Transfer Learning, Transformers, Sequence Models, Image Recognition, Classification, Recurrent Neural Networks (RNNs), Siamese Neural Networks, Artificial Neural Networks (ANN), Convolutional Neural Networks (CNNs), TensorBoard, Generative Adversarial Networks (GANs), Topic Modeling, Linux, Language Models, Python 3, Recommendation Systems, Algorithms, Generative Pre-trained Transformer 3 (GPT-3), XGBoost, Data Extraction, Data Manipulation, Snowflake, Analytics, MySQL, Handwriting Recognition, Google Vision API, Data Analysis, Version Control Systems, Communication, Data Engineering, Regression, RAPIDS, APIs, Amazon SageMaker, NoSQL, Image Search, Amazon Rekognition, AI Design, Jupyter Notebook, Flask, Microservices Architecture, Data Scientist, Chatbots, Jupyter, Cloud, Google Cloud Platform (GCP), Anomaly Detection, OpenAI GPT-4 API, Chatbot Conversation Design, Integration, Programming, AI Programming, Data Cleaning, Unstructured Data Analysis, Machine Learning Automation, Image Analysis, Large Language Models (LLMs), Project Management, Regex, Statistical Analysis, Data Reporting, Data Cleansing, You Only Look Once (YOLO), Elasticsearch, Azure SQL Databases

Software Engineer

2018 - 2019

Yamaha Motor Solutions

Conceptualized and built a ConvLSTM model on TensorFlow for hand gesture recognition. Used Bayesian optimization techniques for hyperparameter tuning.
Worked on single-shot face recognition using SSD MobileNet for detection and Inception-ResNet v2 for feature extraction. Introduced TP-GAN for photorealism and identity, preserving the frontal face-view synthesis from any pose.
Led a team to propose and implement predictive analysis for retail sales forecasting using ARIMA, exponential smoothing, and Holt's Winter models. This resulted in a 5.35% increase in quarterly sales.
Implemented conditional Wasserstein generative adversarial network (cWGANs) to synthesize new vehicle designs from given samples.

Technologies: Python 3, Machine Learning, Data Science, SQL, Artificial Intelligence (AI), Neural Networks, Data Visualization, Optimization, Code Review, Graphics Processing Unit (GPU), Transfer Learning, Machine Vision, Sequence Models, Classification, Siamese Neural Networks, Convolutional Neural Networks (CNNs), Artificial Neural Networks (ANN), TensorBoard, Generative Adversarial Networks (GANs), Linux, Algorithms, XGBoost, Data Extraction, Data Manipulation, Snowflake, Analytics, MySQL, Handwriting Recognition, Data Analysis, Version Control Systems, Communication, Data Engineering, Regression, APIs, Amazon SageMaker, NoSQL, AI Design, Jupyter Notebook, Flask, Data Scientist, Jupyter, Cloud, Integration, Programming, AI Programming, Data Cleaning, Unstructured Data Analysis, Machine Learning Automation, Image Analysis, Project Management, Regex, Statistical Analysis, Data Reporting, Data Cleansing, You Only Look Once (YOLO), Point Cloud Data, Point Clouds, Azure SQL Databases

Experience

AI-based Document Capturing & Processing | AIDCAP

Fully automated the document capture and information extraction process using NLP and computer vision, which mainly dealt with invoices and orders from different sources, including emails, PDFs, DOCX, or images.

I also built and trained custom models for region proposal and text recognition with enhanced semantic understanding using Transformer.

Search Relevance & Ranking

https://www.airtelxstream.in/

Improved search relevance and ranking metrics like click-through rates (CTR), conversion rates, and ad revenue for organic and ad merchants at an eCommerce site.

1. Alternated search keyword generation with semantic search, increasing ad revenue by 20%. The system is scaled to handle 2,000+ requests per second using a Milvus-based vector search and is automatically refreshed daily.

2. I developed a custom deep semantic matching model trained to match search queries with products. I also worked on search query classification using custom-trained embeddings for query understanding to match the search query with product leaf categories.

3. Improved CTR by 50%, leading to a 37% incremental uplift in revenue and a 17% decrease in discount utilization. I improved the CTR using ML-based learn-to-rank models, starting from simple linear regression and tree-based models like LightGBM and XGBoost. These were combined with deep models for search-query understanding.

Callup AI | Chatbot Platform

A generative chatbot for user queries using a Transformer-based Seq2Seq model from chat messages and it is currently deployed commercially, handling thousands of daily conversations in English or Hinglish with almost 85% accuracy.

The chatbot is also comprised of the nearest neighborhood-based intent detection solution, which uses pre-tagged customer messages to tag new customer messages for the support team. The idea was to reduce customer service workload by automatically replying to user messages with pre-defined responses based on intent detection.

Candidate Recommendation Engine

A tool to automate matching job descriptions with resumes, facilitating the process of getting top candidates out of thousands based on several parameters like experience, education, skillset, projects, and location.

By using FastText, NER, LayoutLM, Layout Parser, and a tree-based
embedding search algorithm, we were able to sift through thousands of resumes to find the best fit for the job. We also added new features like face search and real-time blacklisting.

Overall, we reduced manual effort by over 50%.

Retail Sale Forecasting

https://www.yamaha-motor-india.com/

Analyzed the sales trends and sizing curve distribution across categories and product lines to give recommendations on the localized sizing distribution for a demand forecast model. Then, I proposed and implemented predictive analysis for biweekly retail sales forecasting using ARIMA, exponential smoothing, and Holt-Winters methods.

Additionally, I incorporated Bayesian hyperparameter tuning with hybrid modeling using principal component analysis, support vector machines, and ensemble learning methods with Random Forests and XGBoost. This led to a 10% year-on-year increase in key metrics such as sales revenue.

Churn Prediction

Initiated customer churn prediction in a subscription-based purchase model. The first use case was for a telecommunications provider, and the second was for a retail giant, where I leveraged structured and unstructured sales and customer information, including chats, ratings, reviews, emails, and voice chat transcripts for designing KPIs. The model could predict churns almost two weeks before it happened.

I also planned and implemented an A/B test framework and SQL queries. The framework defined key metrics to analyze model performance statistical significance.

Virtual Try-on

A real-time apparel-based virtual try-on tool. It uses a cutting-edge complex ML pipeline that provides a real-time, immersive, clothing-trial experience on a phone or kiosk with a single image and well-thought-out recommendations to help boost sales.

The tool increased in-store purchases by over 25% and via-app purchases by over 20%.

Face Dedupe Platform | Face Recognition, Matching, & Retrieval

A platform with various microservices, including face recognition, face matching, and face liveness-check models.

I updated the platform with face segmentation, face blur detection, document forgery detection, real-time blacklisting, and face search to improve service efficiency and reduce manual effort by over 25%.

Price Optimization

https://tiki.vn/

Optimized retail product prices for Tiki using demand prediction by ML models and time-series historical sales data; LightGBM model is trained to predict sales at daily and 3-day frequencies.

The work included analyzing data, engineering features, managing stakeholders, designing A/B experiments, and performing statistical analysis for hypothesis testing.

Education

2014 - 2018

Bachelor's Degree in Computer Science

Dehradun Institute of Technology - Dehradun, India

Certifications

SEPTEMBER 2019 - PRESENT

Deep Learning Specialization

Coursera

MAY 2018 - PRESENT

Machine Learning Specialization

Coursera

APRIL 2018 - PRESENT

Big Data Modeling and Management Systems

Coursera

MARCH 2018 - PRESENT

Programming the Internet of Things (IoT) Specialization

Coursera

Skills

Libraries/APIs

TensorFlow, PyTorch, Python API, OpenCV, Scikit-learn, Pandas, NumPy, SciPy, Natural Language Toolkit (NLTK), SpaCy, XGBoost, RAPIDS, REST APIs, OpenAI Assistants API, Google Vision API, Keras, Amazon Rekognition, DeepSpeed, OpenAI API, DeepSpeech

Tools

TensorBoard, Amazon SageMaker, ChatGPT, Jupyter, Bitbucket, You Only Look Once (YOLO), Claude, OpenAI Gym, Amazon Elastic Container Service (ECS), Dialogflow, NGINX, GIS, CPLEX, Apache Airflow, AutoCAD, Zapier, GraphRAG, DeepSeek, AI Prompts

Languages

Python, SQL, Python 3, Regex, Bash, Snowflake, JavaScript, Bash Script

Frameworks

Flask, Accelerate, Multi-armed Bandits (MABs), LlamaIndex

Paradigms

Siamese Neural Networks, ETL, Microservices Architecture, Automation, Anomaly Detection, Microservices, Model Context Protocol (MCP)

Platforms

Linux, Jupyter Notebook, Docker, Azure, Amazon Web Services (AWS), Google Cloud Platform (GCP), Kubernetes, Mixpanel, Azure AI Studio, Databricks, Blockchain, Voiceflow, Ollama

Industry Expertise

Project Management

Storage

Azure SQL Databases, PostgreSQL, MySQL, NoSQL, Google Cloud, Elasticsearch, Neo4j, Graph Databases

Other

Machine Learning, Deep Learning, Natural Language Processing (NLP), Computer Vision, Software Deployment, Software Design, Statistics, Statistical Modeling, Data Analytics, Machine Learning Operations (MLOps), Image Processing, BERT, Long Short-term Memory (LSTM), Data Science, Predictive Modeling, Hugging Face, Artificial Intelligence (AI), Optical Character Recognition (OCR), Neural Networks, Recurrent Neural Networks (RNNs), Convolutional Neural Networks (CNNs), Artificial Neural Networks (ANN), Time Series Analysis, Entity Extraction, Classification, Image Recognition, Image Retrieval, Object Detection, Sequence Models, Machine Vision, Transformers, Transfer Learning, Graphics Processing Unit (GPU), Remote Team Leadership, Code Review, Generative Adversarial Networks (GANs), Clustering, Topic Modeling, Data Visualization, Mathematics, Optimization, Language Models, Recommendation Systems, Algorithms, Data Collection, Data Extraction, Data Manipulation, Large Data Sets, Analytics, Data Analysis, Models, Version Control Systems, Communication, Modeling, Data Engineering, Regression, APIs, Image Search, AI Design, Fine-tuning, Causal Inference, Data Scientist, Generative Pre-trained Transformers (GPT), Decision Trees, Chatbots, Cloud, OpenAI GPT-4 API, Chatbot Conversation Design, Integration, Programming, AI Programming, Data Cleaning, Unstructured Data Analysis, Machine Learning Automation, Language Learning, Image Analysis, Large Language Models (LLMs), Statistical Analysis, Data Reporting, Data Cleansing, API Integration, Prompt Engineering, Image Generation, Text to Image, Retrieval-augmented Generation (RAG), LangChain, FastAPI, Multi-agent Systems, Real-time Data, Gemini, AI Agents, Model Context Protocol, Minimum Viable Product (MVP), Distributed Systems, Multivariate Statistical Modeling, GPU Computing, Big Data, Web Scraping, Unsupervised Learning, Generative Pre-trained Transformer 3 (GPT-3), A/B Testing, Revenue Optimization, Pricing Strategy, Handwriting Recognition, Stable Diffusion, NVIDIA TensorRT, Software Architecture, OpenAI GPT-3 API, Financial Modeling, Cognitive Behavioral Therapy (CBT), Data Structures, User Journeys, Customer Journey, Amazon API Gateway, Generative Artificial Intelligence (GenAI), Point Cloud Data, Application Packaging, AI Modeling, Quantization, Vector Databases, Speech-to-Text (STT), Multimodal Models, Internet of Things (IoT), Industrial Internet of Things (IIoT), Computational Biology, Oncology & Cancer Treatment, Genomics, Molecular Biology, Biology, Bittensor, 3D, Diffusion Models, ControlNet, Data Strategy, Data, User Interface (UI), Airtable, Web Applications, Natural Language Understanding (NLU), Point Clouds, Multimodal GenAI, Integrated Development Environments (IDE), Text to Image AI, DALL-E, OpenAI, Mistral AI, Llama 2, Llama 3, Neural Network Pruning, Meta Llama, FSDP, Parallelism, Amazon Bedrock, NVIDIA TensorRT-LLM, Graphics, Image to Video, Time Series Forecasting, Learning to Rank (LTR), Pinecone, Database Optimization, Architecture, AI Chatbots, AI Integration, Zoho, AI Voice Agents, Conversational AI

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