Sagar (Mattikere) A.

Graph-Based RAG Agent for Secure Data Intelligence (EcoGraph-RAG)

Marburg, Germany

Experience

Oct 2025 - Present
2 months
Marburg, Germany

Graph-Based RAG Agent for Secure Data Intelligence (EcoGraph-RAG)

Philipps University Marburg

  • Designed GraphRAG system combining semantic vectors (Chroma) + knowledge graphs (NetworkX/Neo4j) for multi-hop Q&A on climate policy documents.
  • Deployed Llama 3/Gemma via Ollama for $0-cost local inference; achieved ~95% entity-relation extraction accuracy.
  • Built ingestion pipeline for PDFs and 48k-row CSVs; applied grouped median imputation and fixed data sparsity.
Jul 2025 - Aug 2025
2 months

Twitter Sentiment Extraction with BERT + Explainability

  • Fine-tuned BERT-base on 50k+ tweets for multi-class sentiment (85% F1); deployed via Hugging Face Transformers.
  • Integrated LIME/SHAP to generate human-readable feature attributions; visualised in Streamlit for stakeholder review.
May 2024 - Jun 2025
1 year 2 months

EV Insights ML API with Docker

  • Deployed production-ready Flask API in Docker; predicted EV range with Random Forest.
  • Built ETL pipeline; served real-time insights to the dashboard.
Oct 2022 - Jun 2023
9 months
Bengaluru, India

Research Assistant

Dayananda Sagar University

  • Coordinated comparative analysis of algorithms (SVM, Logistic Regression, Random Forest) to optimise diagnosis; achieved model evaluation of 96.73% using SVM models on reference datasets.
  • Applied Vgg16 transfer learning for feature extraction and optimised deep learning model performance on limited medical imaging data to distinguish between benign and malignant tumours.
  • Assisted in synthesising research outcomes and contributed to manuscript drafting using LaTeX and Microsoft Office, supporting publication for ICRTC-2023.
  • Engineered data pipelines using Python and TensorFlow over a 6-month project to automate preprocessing and augmentation of 10,000+ medical imaging records, reducing manual data handling by 70% and accelerating downstream model training cycles.

Summary

Master's student specialising in AI/ML with expertise in ML, RAG systems and LLM deployment.

Proven track record developing production-ready AI solutions for threat detection and secure data intelligence applications.

Languages

English
Advanced
German
Elementary

Education

Apr 2024 - Present

Philipps University Marburg

Master of Science in Computer Science · Computer Science · Marburg, Germany

Oct 2019 - Jun 2023

Dayananda Sagar University

Bachelor of Technology in Computer Science · Computer Science · India

Certifications & licenses

British Airways- Data Science Job Simulation.

Forage

Need a freelancer? Find your match in seconds.
Try FRATCH GPT
More actions