George Mason University
M.S. in Data Analytics Engineering
CGPA 3.87
Machine LearningBig Data AnalyticsData MiningDatabase SystemsStatistical Analysis (R)
May 2024
May 2026
Data withclarity,AI with intent.
Eshwar Anish Varma
Gen AI Data Engineer
Gen AI · Analytics · Engineering
I build audit-traceable AI and analytics systems that turn complex information into useful, human decisions. My work combines OCR, RAG, LLM evaluation, and data engineering to create reliable workflows for regulated documents, ERP automation, predictive analytics, and real-time reporting. I focus on systems that remain explainable, measurable, and practical for the people using them.
ExploreSelected accomplishments
60%
less manual review for regulatory drafting
80%
less daily ERP operating work through automation
200+
live business metrics processed
1st runner-up
SAP Hack2Build Hackathon
01Education
The foundations.
Formal training in data analytics and computer engineering, strengthened by applied research and project work.
Manipal International University
Bachelor of Computer Engineering
2019
2023
02Selected experience
Building useful systems, end to end.
A mix of data engineering, applied AI, teaching, and product-minded analysis.
Graduate Teaching Assistant
George Mason University
Supported instruction for 100 students in Python and Java across algorithms, data structures, and statistical problem-solving.
December 2025
May 2026
Gen AI Data Engineer
AI-Assisted FDA 483 & EIR Platform
Built a human-in-the-loop NLP pipeline using OCR, RAG, and LLMs over regulatory rulebooks to auto-draft FDA 483 observations and EIRs.
January 2026
May 2026
Data Analyst
Morr Computing (Morrai – ERP Automation & Analytics AI)
Built Python and SQL data pipelines integrating live ERP, CRM, and e-commerce sources through REST APIs, powering real-time processing of 200+ business metrics.
July 2023
July 2024
Data Analyst Intern
Morr Computing (Morrai – ERP Automation & Analytics AI)
Wrote SQL queries across live SAP HANA ERP tables to collect and label training data for the ERP AI model.
July 2022
July 2023
03Selected projects
Practical work for complex problems.
Systems built to retrieve, predict, automate, and explain.
2025View detailsCloseData engineering & predictive analytics
IPL Data Analytics System
Built an end-to-end PySpark and Databricks pipeline processing 150,000+ match records, applying EDA and statistical analysis to surface match, player, and venue insights. Engineered predictive features reaching about 70% match-outcome prediction accuracy, and designed ETL workflows and Tableau dashboards for non-technical audiences.
Data engineering & predictive analytics
IPL Data Analytics System
Built an end-to-end PySpark and Databricks pipeline processing 150,000+ match records, applying EDA and statistical analysis to surface match, player, and venue insights. Engineered predictive features reaching about 70% match-outcome prediction accuracy, and designed ETL workflows and Tableau dashboards for non-technical audiences.
What I did
- Designed an end-to-end PySpark and Databricks pipeline for more than 150,000 IPL match records.
- Performed exploratory and statistical analysis across player, venue, and match-level data.
- Engineered predictive features and translated the results into Tableau dashboards for non-technical users.
Outcome
Reached about 70% match-outcome prediction accuracy.
PySparkDatabricksEDATableauStatistics
2025View detailsCloseMachine learning & product delivery
MovieLens Recommendation Engine
Built a collaborative-filtering recommendation system on 100,000+ ratings, tuning models to an RMSE of 0.85 with scikit-learn, cross-validation, and hyperparameter tuning. Deployed an interactive Streamlit application serving personalized recommendations and predicted ratings.
Machine learning & product delivery
MovieLens Recommendation Engine
Built a collaborative-filtering recommendation system on 100,000+ ratings, tuning models to an RMSE of 0.85 with scikit-learn, cross-validation, and hyperparameter tuning. Deployed an interactive Streamlit application serving personalized recommendations and predicted ratings.
What I did
- Prepared and modeled more than 100,000 MovieLens ratings using collaborative filtering.
- Used cross-validation and hyperparameter tuning to compare and improve recommendation quality.
- Built a Streamlit interface that returns personalized recommendations and predicted ratings.
Outcome
Tuned the recommendation model to an RMSE of 0.85.
Pythonscikit-learnCollaborative FilteringCross-ValidationStreamlit
2026View detailsCloseProduct engineering & workflow automation
PicSpace – Photography Client Management Platform
Built a client-management platform integrating email, Telegram, and WhatsApp via REST APIs, automating billing and shipping dashboards for leads, bookings, invoices, and payments. Modeled and deployed on PostgreSQL (Supabase) with Docker-containerized services.
Product engineering & workflow automation
PicSpace – Photography Client Management Platform
Built a client-management platform integrating email, Telegram, and WhatsApp via REST APIs, automating billing and shipping dashboards for leads, bookings, invoices, and payments. Modeled and deployed on PostgreSQL (Supabase) with Docker-containerized services.
What I did
- Built a unified client-management workflow for leads, bookings, invoices, payments, billing, and shipping.
- Integrated email, Telegram, and WhatsApp communication through REST APIs.
- Modeled the application data in PostgreSQL on Supabase and containerized its services with Docker.
Outcome
Created one operational view across the full photography client lifecycle.
PythonREST APIsPostgreSQLSupabaseDocker
2023View detailsCloseComputer vision & automation
Facial Recognition Attendance System
Built a computer vision attendance system in Python and OpenCV, matching faces against an enrolled-user database to automate roll-call.
Computer vision & automation
Facial Recognition Attendance System
Built a computer vision attendance system in Python and OpenCV, matching faces against an enrolled-user database to automate roll-call.
What I did
- Developed the recognition workflow in Python with OpenCV.
- Matched detected faces against a database of enrolled users.
- Used successful matches to automate attendance capture and replace manual roll-call.
Outcome
Turned face recognition into a practical attendance workflow.
PythonOpenCVComputer Vision
04Professional credentials
Learning that compounds.
Verified foundations across cloud, APIs, and applied data work—kept practical and connected to what I build.
AWS Academy
AWS Academy Graduate: Cloud Foundations
Completed foundational AWS training covering core cloud concepts, key services, security, and architecture.
View credential
Coursera
APIs by Meta
Authentication, REST APIs, ordering, debugging, authorization, filtering, serializers, and deserializers.
View credential
Foundations: Data, Data, Everywhere
Data cleaning, data analysis, SQL, spreadsheets, and data visualization.
View credential05Awards & recognition
06Capabilities
The toolkit.
PythonSQLRJavaPySparkPandasNumPySciPyDatabricksLLMsRAGNLPEmbeddingsOCRLLM EvaluationMachine LearningComputer VisionFastAPIStreamlitDockerREST APIsPostgreSQLSupabaseSAP HANATableauPower BIMatplotlibSeabornPlotlyJupyterGitWorkflow Automation
Follow the work.
Start a conversation.
Explore the code behind my projects or connect with me professionally.
Find me on
GitHub
Find me on
LinkedIn