Prem Ganesh Maddirala
Software Engineer - AI/ML | LLMs · RAG · NLP | AI/ML @ Leaniar | MS Data Science @ ASU (May 2026) | Open to Full-Time
- Role
- Software Engineer - Ai Ml at Leaniar
- Location
- Tempe, AZ, US
- LinkedIn followers
- 500 followers
About Prem Ganesh Maddirala
I build production-grade AI and ML systems that solve real business problems.Currently at Leaniar, I\'m engineering an LLM-powered SAP automation platform using RAG, prompt engineering, and OData APIs — cutting processing time from 8 hours to 4 minutes and processing 1,800+ rows at 99%+ accuracy.My work spans the full ML stack: from fine-tuning BERT on 1.6M tweets (92.7% accuracy) and building knowledge graphs in Neo4j, to deploying real-time Kafka streaming pipelines on Kubernetes.What I specialize in: LLMs, RAG & Agentic AI (LangChain, LangGraph, CrewAI) NLP & Deep Learning (BERT, Llama, HuggingFace, PyTorch) Data Engineering & Pipelines (Kafka, ETL, Docker, AWS) End-to-End ML Deployment (FastAPI, Flask, React) Statistical Modeling & Predictive AnalyticsMS Data Science @ ASU (GPA: 3.94 | May 2026)Available June 2026 · STEM OPT Eligible · Future H1BOpen to: ML Engineer · Data Scientist · AI Engineer · NLP Engineer US-based github.com/premtheganeshLet\'s connect if you\'re building something interesting with AI.
Experience
Software Engineer - Ai Ml
Sep 2025 — Present · Irvine, CA, US
Built an automated SAP data extraction system using LLMs to dynamically identify optimal OData API endpoints and critical fields across document types, reducing manual preparation from 8 hours to 4 minutes- Optimizing LLM-powered SAP/JDE automation pipeline with RAG-based prompt routing, response caching, and selective retry logic, reducing API calls by ~50% and improving script generation accuracy by ~30%- Architecting LLM integration layers with bidirectional notation conversion and prompt engineering that intelligently analyzes Excel templates to identify critical fields and generate optimal API queries for any SAP document type- Developing automated data validation pipelines processing 1,800+ row datasets daily with 99%+ accuracy, applying statistical quality checks to ensure data integrity and enable real-time business insights.
Education
Arizona State University
Master of Science - MS, Data Science and Analytics
Birla Institute Of Technology and Science, Pilani Dubai
Bachelor of Engineering - BE, Computer Science
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.