H
Harshith Reddy
About
Detail
Denton, Texas, United States
Contact Harshith regarding:
groups
Networking
Résumé
Jobs
verified_user
0% verified
Data Scientist / AI EngineerDatavantAug 2023 - Current (3 years 2 months)• Architected enterprise RAG platform serving 200K+ documents with sub-millisecond P99 latency by designing
hybrid search architecture on Qdrant with payload indexing, query caching, and graceful degradation patterns.
• Designed multi-stage semantic ranking pipeline (BM25, Gemini embeddings, cross-encoder reranking),
improving retrieval precision by 30% as measured by human annotation and LLM-as-judge evaluation.
• Built production ETL system processing unstructured mortgage documents at scale using LlamaIndex and GPT-4;
implemented retry logic, dead-letter queues, and daily monitoring via Cloud Scheduler batch jobs.
• Developed GenAI-driven conversational assistants using LLMs (GPT-4, Claude, and Falcon) integ- Data ScientistTata Consultancy Services (TCS)Jun 2021 - Jan 2023 (1 year 8 months)• Designed and implemented end-to-end data science workflows including data ingestion, cleaning, feature
engineering, modelling, and deployment using Python.
• Conducted exploratory data analysis (EDA) on large, structured datasets using Pandas, NumPy, Matplotlib, and
Seaborn to uncover trends and patterns.
• Built and optimized machine learning models such as Linear & Logistic Regression, Decision Trees, Random
Forest, XGBoost and K-Means clustering using scikit-learn.
• Applied statistical analysis and hypothesis testing to validate business assumptions and model outcomes.
• Developed time series forecasting models using ARIMA/SARIMA and Prophet for demand and performance
prediction.
• Implemen - GData ScientistGrepthor Software Solutions Pvt LtdDec 2020 - Apr 2021 (5 months)• Translated business use case into predictive analytics frameworks. Validated the model by building POC for
predicting asset performance using python for a $16M project.
• Developed and deployed interactive Power BI dashboards and KPI scorecards for predictive analytics, resulting in
a 30% boost in operational efficiency and compliance score.
• Collaborated with DevOps engineers to implement containerized training and batch inference pipelines on ECS
with version-controlled deployment.
• Partnered with business stakeholders and risk teams to identify AI and analytics-driven use cases, aligning
datascience initiatives with organizational KPIs.
• Improved data pipeline efficiency by 40% and enhanced re
Education
verified_user
0% verified
- Masters in Computer ScienceUniversity of North TexasJan 2023 - Dec 2024 (2 years)GPA: 3.77/4
- vBachelors in Computer Scienceveltech universityJun 2017 - Apr 2021 (3 years 11 months)GPA: 7.7/10