Vinod Reddy
Verified Expert in Engineering
Data Science Developer
Bengaluru, Karnataka, India
Toptal member since February 5, 2019
Vinod is a senior ML engineer with 12+ years of experience delivering production AI systems across sports tech, ad tech, and enterprise productivity. He's led the end-to-end development of scalable solutions for forecasting, causal inference, RL, and LLM-based copilots. Lately, he's focused on designing agentic AI workflows and integrating LLMs into business-critical systems. With a strong track record at Dream11 and Rhythms, Vinod blends deep technical expertise with a focus on business impact.
Portfolio
Experience
- Data Science - 6 years
- Natural Language Processing (NLP) - 3 years
- Machine Learning - 3 years
- Generative Pre-trained Transformers (GPT) - 3 years
- Python 3 - 3 years
- Scikit-learn - 2 years
- PyTorch - 2 years
- PySpark - 2 years
Preferred Environment
Python, Git, Linux, MacOS, Tmux, Sublime Text
The most amazing...
...thing I led was the development of a real-time contest engine at Dream11 that used forecasting and policy systems to auto-launch contests.
Work Experience
Principal Data Scientist
Rhythms
- Developed an enterprise-grade agentic chatbot that empowers users to seamlessly navigate and manage their OKR workflows.
- Designed and deployed a Retrieval-Augmented Generation (RAG) agent that intelligently answers user queries related to their OKRs.
- Designed and deployed a prompt structuring and evaluation framework based on LangChain and LangSmith, which helped optimize prompts across the company.
Principal Data Scientist
Dream11
- Built and improved a contest generation engine using forecasting and control systems, which handled generating contests automatically based on real-time data and accounted for 50% of the company's revenue.
- Introduced and pioneered causal methods and algorithms to improve algorithmic and product decisions across the company.
- Built a causal inference platform for in-house use cases and published a paper on the same (Arxiv.org/abs/2401.11932).
- Managed a team of more than 15 data scientists for over three years in delivering value across multiple use cases involving forecasting and control systems, causal inference, personalization and recommendations, and Generative AI (GenAI).
- Built and deployed end-to-end voice AI suggestions that help users in creating teams for contests.
Data Scientist
BryghtAI
- Built a diagnostics system for chatbots which calculates the performance of the bot in terms of metrics like navigational efficiency, customer satisfaction, etc.
- Built models to extract intent, entities, sentiment, abuse from user-bot chat corpus. Compared performance of NLU systems like Deeppavlov, Rasa NLU, and Snips NLU on the custom chat corpus.
- From the extracted features, trained gradient boosting machines to predict custom performance metrics.
- Built short text topic models to figure out topics in an unlabeled chat corpus.
Data Scientist
Thirdwatch AI
- Built an eCommerce RTO/fraud detection system to predict whether an order is returned back without being delivered.
- Using various ML techniques, extracted informative features from transactional data of over 70 million rows. Combined information from several sources of data like user history, user personal data, clickstream data, geographic address data, item RTO history, and item information.
- Used RNN-based models to parse text addresses into individual components and used these components to extract informative features.
- Built predictive models from the extracted features for the RTO prediction. Compared performance across gradient boosting machines, random forests, and neural network models using metrics like precision, recall, and F1 score.
Data Scientist
Money Control
- Built a system which gives pointed answers for natural language questions from a corpus of budget article data. Used a pipeline of TfIdf-based retrieval and glove vector based recurrent neural network system to answer the queries.
- Optimized the pipeline to decrease the latency and increase the throughput to t the production requirements of MoneyControl. Achieved a throughput of 100 requests/second with a latency of one second.
- Improved the performance of pipeline by incorporating Solr instead of the T df-based retriever and text summarizer to better show the results. Achieved a top accuracy of 75% on test data.
Data Scientist
2020Imaging
- Implemented a YOLO-based people counter system to count the number of people in the camera view and detect crowds. Trained the darknet YOLO system on a custom dataset to improve accuracy.
- Trained and implemented a face attention network on a custom dataset to detect masked faces in the view of the security cameras.
Co-founder/Tech Lead
Rocketbox
- Built an Uber-like system from the ground up for the logistics domain.
- Designed and developed registration and billing for driver and customer, duty and attendance management, vehicle tracking, order management, ad-hoc driver management, role-based admin, and client modules.
- Developed client and admin web portals as well as client and driver Android applications.
- Managed a team of developers and interacted with the operations team to gather requirements, create and assign tasks, manage sprints, and review and evaluate performance.
- Built and maintained software using Chef and Capistrano for deploying code and spinning up compute and database servers.
- Integrated and managed cloud infrastructure on AWS, EC2, RDS, and MongoAtlas.
- Integrated various services like Freshdesk for managing complaints, Kookoo for managing SMS, Mapbox for rendering maps, Google Maps API for improving the vehicle tracking data and distance calculations, Pusher for web notifications, GCM notifications for Android, and more.
Senior Software Engineer
Directi/media.net
- Contributed to designing and rewriting the ad-serving system from PHP to Java8. Proposed and educated colleagues in the usage of Java 8 for re-write.
- Built a system to optimize keywords in the ad-serving to include per country factor.
- Incorporated the country factor in different keyword extraction flows.
- Built internal tools for developers to view statistics and keyword performance metrics based on different parameters.
- Planned and implemented education of peers in ML technologies.
- Presented various ML algorithms like LDA, SVM, bagging, boosting, and GBM as well as database technologies like MongoDB, CouchDB, Memcache, and Redis.
Experience
QA Chatbot
Chatbot Diagnostics
People Counting in Security Cameras
Face Mask Detection in Security Cameras
RTO/Order Returns Prediction in eCommerce
Education
Bachelor of Technology Degree in Computer Science and Engineering
Indian Institute of Technology Bombay - Mumbai, India
Certifications
Deep Learning Specialization
Coursera
Skills
Libraries/APIs
XGBoost, PySpark, Scikit-learn, PyTorch, TensorFlow Deep Learning Library (TFLearn), TensorFlow, Pandas, Pydantic
Tools
H2O AutoML, Amazon SageMaker, AI Prompts, Claude Code, Sublime Text, Tmux, Git, Chef, Docker Compose, Docker Swarm, GraphRAG, Grafana
Languages
Python, Python 3, SQL, JavaScript, Ruby, Java, PHP, C++, Snowflake, TypeScript
Frameworks
Spark, LightGBM, Ruby on Rails 5, Ruby on Rails (RoR), Caffe, LangGraph, Ruby on Rails 6, DSPy
Platforms
Amazon Web Services (AWS), Databricks, H2O Deep Learning Platform, Docker, Harness, MacOS, Linux, NVIDIA CUDA, Kubernetes, Azure
Storage
MongoDB, PostgreSQL
Paradigms
Anomaly Detection, Model Context Protocol (MCP)
Other
Data Science, Machine Learning, Deep Learning, Chatbots, LangChain, Retrieval-augmented Generation (RAG), Large Language Models (LLMs), Artificial Intelligence (AI), AI Algorithms, API Integration, Prompt Engineering, AI Chatbots, Forecasting, Predictive Modeling, Predictive Analytics, Machine Learning Operations (MLOps), Large Data Sets, Data Modeling, ML Pipelines, Natural Language Processing (NLP), Algorithms, Generative Pre-trained Transformers (GPT), Agentic AI, AI Agents, Generative Artificial Intelligence (GenAI), Chatbot Conversation Design, Conversational AI, Time Series Forecasting, Vector Databases, Vector Search, Fine-tuning, Recommendation Systems, FastAPI, Multi-agent Systems, Stock Trading, Stock Market, Charts, A/B Testing, ETL Pipelines, Model Validation, Statistical Modeling, CI/CD Pipelines, Classification, Data Analysis, AI Agent Orchestration, Light LLMs, Pgvector, RAG Pipelines, Model Evaluation, Pricing, Pricing Models, AI Programming, AI Design, Computer Vision, Image Processing, Causal AI, Causal Inference, Generative Systems, Large Language Model Operations (LLMOps), Computer Science, OpenTelemetry, ElevenLabs Solutions, Knowledge Graphs, Regression Testing, Full-stack
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
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring