Dowtya Sri Prasanth Vemula

6145304641
[email protected]

EDUCATION:
 Master of Science in Artificial Intelligence and Machine Learning - University of North Texas
Relevant Courses: Machine Learning, Deep Learning, Feature Engineering, Data Visualization, Software
Development for AI, Reinforcement Learning using Neural Networks, Methods in Empirical Analysis, Statistics,
Data Structures and Algorithms and Problem Solving.
 Bachelor of Technology in Computer Science and Engineering - Amrita Vishwa Vidyapeetham
 Relevant Courses: Software Engineering, Database Management Systems, Data Structures and Algorithms, Big
Data Analytics, System Design
SKILLS:
 Machine Learning & Data Science: Transfer Learning, Model Evaluation (accuracy, precision, recall), API
Development
 Programming Languages: Python, SQL, Java, Bash, JavaScript, Scala, C, C++, C#, R, CUDA
 Cloud Platforms & APIs: AWS (EC2, S3, Lambda, RDS, SageMaker, Bedrock, CloudWatch, Glue, Redshift), Azure
(OpenAI, Cognitive Services, Machine Learning, AKS, Functions, DevOps, AutoML), GCP (Vertex AI, Cloud
Functions, Cloud Storage, BigQuery, Cloud Run, Cloud AI Platform, Document AI, Cloud Vision API, Rest Api
Design)
 Computer Vision: Object Detection (YOLO, Faster R-CNN), Image Segmentation (U-Net, Mask R-CNN, CLIPSeg),
Image Classification (ResNet, Inception), Feature Extraction, Image Augmentation, Transfer Learning
 Frameworks & Packages: TensorFlow, Pandas, NumPy, scikit-learn, Matplotlib, Seaborn, Keras, SciPy, OpenCV,
StatsModels, PySpark, PyTorch, NLP (BERT, openNER, spaCy, GLiNER, openNLP), LangChain
 Machine Learning & Statistics: Regression, Classification, SVM, Decision Trees, Random Forests, Clustering,
Gradient Boosted Trees, Neural Networks (LSTM, GANs, CNN, Transfer Learning), PCA, K-NN, XGBoost
 MLOps Tools: Docker, Kubernetes, Terraform , Ansible, Jenkins, Grafana, FastAPI, MLflow, Ray, Databricks,
Snowflake, Azure DevOps
 Generative AI & Vector Databases: GANs, VANs, Diffusion, RAG, LLMs, Transformers, Prompt Engineering,
Knowledge Graphs (Neo4j), Vector Databases (Pinecone, Weaviate, Milvus, Vespa, Chroma)
PROFESSIONAL EXPERIENCE:
Senior Ai Engineer -Benchmark Gensuite Nov 2024 – Present
Responsibilities:
 Optimized supervised and unsupervised Machine Learning workflows by designing a Platform Agent with LangGraph and
LangChain, processing terabytes of data on Apache Spark and Hadoop HDFS, reducing manual query handling by 35% and
accelerating model training pipelines.
 Engineered scalable big data pipelines by implementing Apache Spark jobs integrated with Teradata for data
warehousing, transforming raw datasets into feature sets for ML models, boosting data processing throughput by 50%.
 Deployed production ML models by containerizing inference services with Docker and Kubernetes and exposing APIs via
FastAPI and Python, achieving 99.8% uptime and reducing client-reported issues by 25%.
 Developed advanced embedding pipelines by leveraging AWS Bedrock Titan v2 and Cohere multilingual models alongside
Apache Spark and Hadoop, automating data chunking and model routing, improving text-processing throughput by 50%
and semantic search relevance by 45%.
 Refactored risk-classification logic by centralizing rule sets in PySpark and Teradata SQL, standardizing checks across 68
risk types, reducing regression bug reports by 60% and ensuring consistency in production.
 Architected a dual-agent document-processing system by orchestrating Hadoop-based ETL pipelines with AWS Textract
for OCR and OpenSearch Serverless for embeddings, reducing LLM call volume by 55% and cutting inference costs by
40%.
 Monitored and analyzed production ML model performance by querying Spark SQL and Pandas-based data analysis
pipelines, detecting anomalies in production logs, and automating re-training triggers to maintain 98% model accuracy.
 Presented machine learning solutions and big data insights to clients by developing interactive dashboards in AWS
CloudWatch and Tableau, enhancing stakeholder understanding and driving a 25% improvement in decision-making
speed.
 Led knowledge-transfer workshops on Hadoop, Apache Spark, and Teradata for cross-functional teams, mentoring new
engineers to accelerate skill acquisition by 70% and promote best practices in ML and big data processing.
 Streamlined CI/CD pipelines for ML model deployment by configuring Jenkins and GitLab to automate Docker builds,
tests, and Kubernetes rollouts, reducing deployment time from 3 hours to 45 minutes and ensuring reproducibility.

Environment: Python, Django, Pandas, Linux/Ubuntu, Machine learning, Artificial Intelligence, MLLib, NLP, Azure, RabbitMQ,
Restful web service, Shell Scripting, Docker, Bit Bucket, Github, CI/CD Pipelines, MySQL, Flask API, MySQL, RabbitMQ, Azure,
Tensorflow, Keras, Pandas, PyTest Responsible for systems implementation and business analysis related to streamlining
operations for ALM operations, especially related to mortgage-based products and their derivatives.
Generative Data Engineer - Salesforce Dec 2023 – Nov 2024
Responsibilities:
 Integrated large-scale Salesforce datasets into AWS S3 by deploying MuleApps in AnyPoint, automating data
ingestion pipelines with Apache Spark on Hadoop and Teradata connectors, reducing manual entry by 50% and
improving data analysis accuracy.
 Automated CPS property data generation by orchestrating Airflow DAGs with YAML configurations and Ninja
templates, loading terabytes of raw data from S3 into Snowflake and Teradata, increasing processing speed by
30% and cutting errors by 90%.
 Designed and deployed RAG pipelines for GenAI chatbots by integrating LangChain, AWS Bedrock models, and
OpenSearch Serverless on Hadoop clusters, improving contextual relevance by 35% and enabling scalable ML
model inference.
 Engineered a dynamic Router Query Engine using PySpark and Teradata SQL for big data routing logic, calling
specialized modules to process ML tasks, reducing query misrouting by 25% and enhancing system reliability.
 Developed data persistence strategies in AWS S3 with Snowflake and Teradata for streaming contexts, utilizing
Spark Structured Streaming on Hadoop to support real-time ML model updates with low-latency data availability.
 Built production-ready JavaScript RAG web apps by leveraging Node.js, React, and Elasticsearch on Hadoop HDFS,
delivering conversational ML interfaces with 35% higher user engagement.
Environment: Python, FastAPI, Streamlit, LangChain, LangGraph, AWS Bedrock (Titan v2, Cohere Multilingual, Claude via
RAG), AWS OpenSearch Serverless/Elasticsearch, AWS Textract, Mistral OCR, AWS Cognito, AWS API Gateway, Amazon S3,
KNN/ANN algorithms, AWS CloudWatch
AI Engineer Intern - Amity Tech Corporation Sep 2023 - Dec 2023
 Engineered AI chatbot leveraging GPT-3.5 and LangChain with Streamlit, deploying models on Hugging Face Spaces and
Docker, cutting client response times by 40% and boosting satisfaction by 35%.
 Optimized semantic search pipelines by building embedding-based retrieval using Cohere LLMs and AnnoyIndex on
Apache Spark, achieving a 25% increase in retrieval accuracy and reducing query latency by 20%.
 Developed end-to-end ML pipelines in Python and Hadoop for document ingestion, vector indexing, and query
processing, enabling scalable AI-driven support solutions with 50% faster resolution.
 Built interactive dashboards with Pandas, NumPy, and matplotlib in Streamlit, presenting ML performance metrics and
data analysis insights to stakeholders, enhancing transparency.
 Collaborated with cross-functional teams to integrate Python-based Machine Learning models with production-grade
RESTful APIs, supporting deployment on cloud platforms with 99% uptime.

Environment: Python, OpenAI GPT-3.5 API, LangChain, Streamlit, Hugging Face Spaces, Cohere LLMs, AnnoyIndex, semantic-
search & embedding pipelines, Python data-processing libraries (Pandas, NumPy), machine-learning frameworks (scikit-
learn), interactive dashboarding tools.
Graduate Teaching Assistant - Dept. of CSE, University of North Texas, Denton, TX. Aug 2022 – Sep 2023
 Tutored 200+ graduate and undergraduate students in statistics, probability, and advanced computer science
principles, incorporating machine learning concepts into lessons, which boosted average grades by 15% and
enhanced student comprehension.
 Led a wildfire spread prediction project by applying meteorological and topographical data to machine learning
models, improving next-day prediction accuracy by 20% through feature engineering and model tuning using
Python and scikit-learn.
 Developed and optimized Python-based workflows for satellite image processing in a geo-spatial lab, utilizing QIS,
GDAL, and NumPy, increasing image processing efficiency by 25% and enabling real-time analysis of large
datasets.
 Incorporated machine learning techniques into geospatial data visualization by leveraging Python libraries
(Matplotlib, Plotly) to create interactive dashboards, enabling students to visualize complex data sets and
increasing their analytical capabilities by 30%.
 Mentored students on research methodologies and project development in computational science, guiding them
in applying advanced algorithms and statistical models, which led to the publication of 2 student research papers
in peer-reviewed journals.
 Collaborated with faculty and industry professionals to incorporate the latest advancements in machine learning
and geospatial analytics, elevating the quality of teaching materials and project assignments.
 Implemented practical, hands-on assignments for students, introducing Python’s TensorFlow and Keras
frameworks for real-world data modeling, enhancing their exposure to state-of-the-art machine learning
techniques by 40%.
Site Reliability Engineer Intern - Honeywell, Bangalore, India Jan 2022 –
Jul 2022
 Developed JavaScript-based RAG web applications with Node.js, Express, and React, integrating Elasticsearch
on Hadoop for vector retrieval, increasing user engagement by 35%.
 Built retriever engines using LangChain.js and Redis for conversational memory, optimizing big data query
processing on Teradata and Hadoop clusters for 25% faster response.
 Implemented Docker and Kubernetes-based CI/CD pipelines in GitLab for continuous deployment of ML
inference services, reducing release cycle time by 40%.
Environment: JavaScript (ES6+), Node.js, Express.js, React.js, LangChain.js/RAG frameworks, Elasticsearch / OpenSearch
(vector retrieval), Redis (conversation memory), Docker, Kubernetes, Git & CI/CD pipelines.
Teaching Assistant - Dept. of CSE, Amrita Vishwa Vidyapeetham, Coimbatore, India. Jun 2018 – Dec 2021
 Illustrated fundamentals of supervised and unsupervised Machine Learning by mapping Scratch and Flogarithm
activities to K-means clustering and decision-tree workflows in visual programming environments, enhancing
student mastery of ML basics by 45%.
 Designed interactive lab sessions simulating big-data processing pipelines with pseudo–Apache Spark jobs
implemented in Python, demonstrating parallel data transformations and boosting engagement and proficiency
in programming basics by 30%.
 Mentored cohorts of 50+ undergraduates in C, Python, Data Structures, and Algorithms through hands-on coding
exercises and Teradata-style SQL query workshops, improving exam pass rates by 25% and solidifying
foundational data-analysis skills.
 Developed Python-based algorithmic assignment templates and leveraged a simulated Hadoop HDFS
environment to teach distributed file storage and map-reduce concepts, increasing student confidence in Big
Data workflows by 40%.
 Facilitated weekly solution walkthroughs and technical presentations on computational thinking and algorithm
 design using structured slide decks and live coding demos, strengthening both peer-to-peer communication and
student feedback scores by 35%
ACADEMIC PROJECT EXPERIENCE
Intelligent Resume Matching Platform May 2024 – Present
 Developed an automated resume screening tool with Streamlit, LangChain, and Pinecone, enhancing
candidate-job match accuracy by 50% and optimizing high-volume HR workflows.
 Built a vectorized resume matching system using Sentence Transformers and Pinecone, reducing candidate
screening time by 40% through efficient, similarity-based retrieval.
Customer Care Call Summary Alert Jun 2024 – Jul 2024
 Developed an AI-driven call summarization and alert system with Streamlit, OpenAI's Whisper, and Zapier,
automating customer service call summaries and delivering real-time alerts to stakeholders, leading to a 30%
reduction in manual reporting time.
 Designed a zero-shot NLP architecture using LangChain and Whisper to transcribe, summarize, and notify
stakeholders with Zapier integration, resulting in improved response times and streamlined communication
workflows.
Invoice Extraction Chatbot Dec 2023 - Jun 2024
 Led the design of an OCR and NLP pipeline using PyPDF and OpenAI models, cutting manual processing time by
50%.
 Created a Streamlit interface for invoice uploads, enhancing user accessibility and processing efficiency by 40%.

AI Powered Support ChatBot For A Website Jun 2023 – Dec 2023

 Developed an AI-powered support chatbot using Streamlit, LangChain, and Pinecone, resulting in a 40%
increase in response accuracy and improved customer query resolution speed.
 Engineered a modular NLP and vector storage architecture with Sentence Transformers and Pinecone, enhancing
user satisfaction by 35% through fast and precise information retrieval.
Automatic Ticket Classification Tool Dec 2022 – Jun 2023
 Created an automated ticket classification system using SVM and Pinecone, reducing manual sorting by 70%.
 Designed a Streamlit interface for efficient ticket processing, enhancing categorization accuracy and scalability.
CSV Data AnalysisTool Dec 2020 – Jun 2021
 Engineered a web-based data analysis app with Streamlit and Pandas, improving insight extraction by 60%.
 Developed NLP-enhanced query processing with LangChain, boosting decision-making efficiency by 40%.