Harshitha

Hi, I'm Harshitha

AI/ML Engineer

I turn data into deployed intelligence by building practical, production-ready AI/ML systems.

About Me

I'm Harshitha. When I'm not building machine learning systems, you'll probably find me chasing mountain views, wildflower seasons, or running through open plains. Utah has my heart, but Boulder will always be my number one. Paragliding in San Diego stands out as one of my most memorable adventures. Skiing is my favorite sport. That same curiosity drives how I approach my work: exploring, learning, and turning ideas into reality.

Harshitha outdoors

Professional Experience

January 2025 - November 2025

Machine Learning Engineer

MODE Civil Services

Built CivAssist, an AI drafting assistant on GCP that automates PDF redline corrections in AutoCAD. Built an end-to-end pipeline combining OCR and NLP with Vertex AI (Gemini) and LangGraph-based agentic system, reducing manual drafting effort by 85% across 1,000+ documents/month.

December 2023 - May 2025

AI Engineer

University of Colorado Boulder

Architected end-to-end Generative AI pipelines on GCP using Vertex AI and LangChain, integrating RAG with knowledge graphs for multi-hop reasoning over institutional research data. Provisioned scalable infrastructure and deployed fine-tuned LLMs to improve forecast reliability and automate research synthesis.

July 2021 - June 2023

Associate Consultant - Data Science

Wipro Technologies

Engineered scalable GCP data pipelines and applied BERT-based NLP models to extract insights from unstructured clinical data, improving patient-outcome prediction models by 15%. Managed the ML lifecycle with MLflow and SHAP, and deployed real-time inference APIs using Docker and Kubernetes.

Education

Master of Science

Data Science

University of Colorado Boulder

August 2023 - May 2025

GPA: 4.0/4.0

Bachelor of Technology

Information Technology

Vellore Institute of Technology

August 2017 - May 2021

GPA: 7.98/10

Skills

ML & AI

Supervised LearningUnsupervised LearningXGBoostRandom ForestLogistic RegressionAnomaly DetectionRecommendation SystemsClusteringComputer VisionOpenCVRAGMCPGoogle ADKSemantic Search

LLMs & NLP

GPT-4gemini 2.0LLaMAMistralLangChainLoRAFAISSNLPPrompt EngineeringAgents

MLOps & Cloud

GCP (Vertex AI, BigQuery, Cloud Storage, Cloud Run, Vector Search, Google Kubernetes Engine (GKE), Pub/Sub, Model Armor)Azure (ML, Synapse, DevOps, App Insights)MLflowDockerKubernetesCI/CDGit

Data Engineering

ETL PipelinesAirflowData PipelinesSAP HANAOraclePostgreSQLDatabricks

Programming & Data

PythonPandasNumPyScikit-learnKerasPyTorchTensorFlowRSQLC++FlaskAPI Design

Analytics & Visualization

TableauPower BILookerStreamlitMatplotlibPlotlyExcelA/B TestingData Storytelling

Projects

Generative AISpring 2025

Buffbot

LLM-Powered Teaching Assistant

Built a document-aware Q&A chatbot using RAG with LangChain, FAISS, and LLaMA. Developed a LangGraph-based agentic workflow and automated pipeline to summarize 200+ lecture videos, reducing manual review effort by 85%.

LangChainLLaMAFAISSStreamlitLangGraph
LLM Application2025

CivAssist

AI Drafting Assistant

Built an AI assistant that automates PDF redline corrections in AutoCAD. Developed the LLM interpretation layer using Mistral models and prompt/schema design to convert handwritten markups into machine-readable edit instructions.

MistralAzureAutoLISPCI/CDPower BI
Full StackFall 2024

LocalCrafts

Workshop Booking Platform

Designed a Docker-containerized Flask-PostgreSQL app with Gemini API for automated content and search. Deployed via Cloud Run with Pub/Sub and maintained consistency using Docker Compose.

FlaskPostgreSQLDockerGCPGemini API
Data ScienceSpring 2024

Unemployment & Crime Analysis

Statistical Correlation Study

Analyzed nationwide unemployment and crime datasets using statistical tests and A/B testing. Designed a Power BI dashboard to visualize trends, significant differences, and data-driven insights.

PythonPower BIStatisticsA/B Testing

Contact

Let's Work Together

I'm always open to discussing new opportunities, interesting ML projects, or just connecting with fellow engineers. Feel free to reach out!