Data · Software · Engineering

Naga Swetha
Vishali Varnajosyula

Data and software engineer with four years of professional experience delivering enterprise data and migration solutions at Accenture across six countries. Now completing an M.S. in Computer Science (4.0 GPA) and building projects in data engineering and machine learning.

01

About

I build reliable systems with data. Over four years at Accenture I delivered end-to-end ETL and enterprise data migration for global clients across six countries — designing data flows, building SQL validation frameworks, and seeing projects through multiple zero-defect production go-lives. As a graduate researcher I engineered an automated pipeline processing over 1.5 billion records. I'm now completing an M.S. in Computer Science (4.0 GPA) and extending that foundation into data engineering and machine learning, applying the same focus on validated, dependable pipelines to new tools and problems.

02

Skills

Enterprise Data / SAP

SAP BODSData migrationETL & data flowsCutover & reconciliationFI / MM / PP / SDClassification (CA-CL)

Languages

SQL (advanced)PythonBashCJava

Data & ML

ETL pipelinesValidation frameworksSchema designPandas / NumPyscikit-learnOpenCV / YOLOv8

Tools & Process

MS SQL ServerJIRA / HP ALMAgile / ScrumGitTableau / Power BILinux / WSL
03

Projects

SAP Master Data Quality & Reconciliation

In Progress

A SQL project modeling core SAP master-data tables (MARA, MAKT, MARC, KNA1) with proper keys, then reconciling a migrated load against the original source — detecting missing records, value mismatches, and data-quality issues, the way a migration team audits a cutover.

SQL · SAP master data · Reconciliation · SQLite · Power BI

SAP Data Migration & Validation Framework

Live

A configuration-driven framework that profiles, cleanses, validates, and reconciles legacy data against a target SAP schema — holding back invalid records and producing a full audit trail. Built on Material Master as the example object.

Python · SQL · Pandas · YAML rules · Data validation

Real-Time Object Detection System

Live

End-to-end YOLOv8 inference pipeline with CLAHE lighting normalization and a threaded capture stage that decouples frame I/O from inference. Sustains ~16 FPS on CPU with stable per-frame latency, visualized with Matplotlib.

YOLOv8 · Python · OpenCV · Matplotlib · Threading · Docker-ready

LLM Methods for Time-Series Forecasting

Live

A comparative study of five LLM-based forecasting approaches — prompting, in-context learning, RAG, and LoRA / AdaLoRA fine-tuning — benchmarked on METR-LA traffic and WeatherBench Z500 datasets, including a documented negative result on retrieval reranking.

PyTorch · Hugging Face PEFT · FAISS · Groq (LLaMA-3.1) · TinyLlama

Real-Time Anomaly Detection Platform

In Progress

An MLOps system scoring streaming transactions in real time: Kafka ingestion, a containerized FastAPI scoring service, Airflow-orchestrated retraining with dbt feature builds, and data-drift monitoring that triggers retraining on distribution shifts.

Kafka · FastAPI · Airflow · dbt · Docker · Cloud
04

Experience

Jan 2025 — Present

Graduate Research Assistant

SIU College of Agricultural Sciences · Carbondale, IL

  • Engineered an automated Bash & Python pipeline processing 1.5 billion+ sequencing records across 472 datasets, with 48-hour unattended execution and structured logging.
  • Increased variant detection 5× (804 → 4,041) by tuning the CRISP variant caller through iterative configuration testing.
  • Built a Python tool parsing 53 MB structured text files into 19 formatted Excel reports with auto-filters and computed aggregations.
Aug 2025 — Dec 2025

Graduate Teaching Assistant — Operating Systems

Southern Illinois University · Carbondale, IL

  • Taught OS concepts (processes, scheduling, memory management) through office hours and review sessions.
  • Built a Python & OpenCV tool to collect and label image datasets, cutting manual labeling effort ~50%.
  • Prototyped LLM-based auto-grading to surface common student error patterns.
Aug 2021 — Aug 2025

SAP BODS Developer / Data Migration Consultant

Accenture · India & France

  • Built SAP BODS jobs, workflows, and dataflows to extract, cleanse, transform, and load 20M+ records from legacy systems into SAP across six countries.
  • Delivered key SAP objects: GL master, Material Master, Batch Master, Bill of Materials, BOM & Material Classification, and Customer Master Relations.
  • Worked extensively with the SAP Classification System (CA-CL) — characteristics, classes, and assignments requiring careful cross-system harmonization.
  • Built SQL validation frameworks and reconciliation dashboards for pre/post-load integrity, reducing post-migration defects ~30%.
  • Led DEV/UAT/PROD deployments and daily stand-ups, delivering multiple zero-defect production cutovers during international go-lives.
2019 — 2020

Research Intern — Railway Signaling & Defense Telemetry

South Central Railways · DRDO · India

  • Contributed to real-time safety and anomaly detection algorithms for distributed railway signaling systems.
  • Applied signal analysis and modeling techniques to defense telemetry data for embedded systems.
05

Education

M.S. Computer Science

Southern Illinois University, Carbondale, IL · Expected May 2027

GPA 4.0 / 4.0

B.E. Electronics & Communication Engineering

SCSVMV University, Tamil Nadu, India · 2017 – 2021

GPA 9.01 / 10
Recognition “Emerging Employee of the Year” at Accenture for innovation in AI & automation · Zero-defect record across international SAP go-lives · State-level athlete (Discus Throw, top-2 ranking).