Dhanush Chandra Shekar

[RAG]

[AGENTIC]

[FINE-TUNED]

System Prompt // v2.0Press / to Search

User

Who is Dhanush Chandra Shekar?

System

AI/ML Engineer who ships, not just experiments.

Initiate Project

View CV

View Case Studies →

Personnel Profile

Originating from the tech hub of Bangalore, I established a rigorous structural foundation in Computer Science, graduating with an 8.51 CGPA before transitioning to advanced machine learning architecture.

Currently, my primary research node is at the Indiana University, Kelley School of Business, pursuing an M.S. in Data Science (3.71 GPA). Here, I orchestrate and deploy end-to-end LLM pipelines utilized daily by faculty for complex data evaluation.

I specialize in combining deep prompt engineering with robust software engineering to deliver reliable, evaluated AI systems at scale. By reducing hallucinations and creating rigorous evaluation benchmarks, I turn chaotic language models into structured, mission-critical assets.

Quick Facts

Education

M.S. Data Science // PyTorch // AWS

Location

Bloomington, IN // Kelley School

Work Ethic

High Efficiency Output

Current Role

Research Asst. Building RAGs

Technical Architecture

The core infrastructure, frameworks, and methodologies utilized across my deployments.

Inference Engine

Neural AI & LLM Core

LangChain
FAISS Vector DB
RAG Pipelines
LLM Fine-tuning
Prompt Engineering
Agentic Workflows

Pattern Recognition

Algorithmic Models

PyTorch
scikit-learn
XGBoost
LightGBM
SentenceTransformers
Model Evaluation

MLOps & Infrastructure

Data Orchestration

Python
FastAPI
Docker
AWS
Neo4j
PostgreSQL
CI/CD Pipelines

Projects & Case Studies

Selected case studies demonstrating production-grade AI implementations and data pipelines.

GraphRAG Financial Knowledge Navigator

PythonNeo4jChromaDB+2

Hybrid RAG pipeline combining Neo4j (3,422 entities) and ChromaDB vector search to ingest 1,605 SEC filings. Achieved superior performance vs baselines with a 46-benchmark evaluation suite.

TelemetrySuperiority via 46-Bench<1.2s Graph Path

RAG-Based LLM System

LangChainFAISSHugging Face+2

Orchestrated LangChain and FAISS vector search system boasting 87% context-aware accuracy. Reduced hallucinations by 65% via LoRA/PEFT fine-tuning on base LLMs.

Telemetry87% Context Accuracy<800ms Retrieval

LodeAI

FastAPIOpenAILangChain+2

Zero-to-one agentic AI platform featuring multi-step LLM orchestration and automated evaluation gates. Deployed to production using Docker, PostgreSQL, and full CI/CD deployment pipelines.

Telemetry100% Eval Pass<2s Agent Route

Demand Forecasting

XGBoostLightGBMMLflow+2

Built robust demand forecasting models benchmarking XGBoost and LightGBM with MLflow experiment tracking (MAE: 0.2575). Monitored for drift using Evidently AI.

Telemetry0.2575 MAE Precision<200ms Inference

InsightFlow

SQLPostgreSQLdbt+2

Automated a multi-source ELT pipeline using Apache Airflow and dbt, reducing data processing time by 65% with zero manual intervention.

Telemetry65% Time ReducedCron Automation

Pricing Optimizer

statsmodelspandasNumPy+2

Developed an OLS regression demand model evaluating 676 datasets to identify profit-maximizing pricing. Implemented interactive Plotly curves.

Telemetry676 Sets EvaluatedReal-time Plots

Experience Log

A chronological record of engineering environments, system integrations, and professional certifications.

Research Assistant – AI & LLM Engineering

Indiana University, Kelley School of Business

June 2024 – Present

Designed and deployed an end-to-end LLM pipeline adopted daily by faculty, combining RAG, information retrieval, and automated evaluation.
Reduced AI output inaccuracies 30–40% by iterating on retrieval configurations and prompt engineering strategies.
Defined KPI thresholds and success metrics for AI model outputs, enabling reproducible benchmarking.

AWS Certified AI Practitioner

Amazon Web Services

2024

Official credential validating overall understanding of AI, ML, and generative AI concepts and use cases on AWS.

SYSTEM_MESSAGE // 0xFINAL

Connection: Secure

[SYS]:End of Dhanush's Archive. The system is standing by to initiate deployment or transmit further documentation. Would you like to proceed?

[Execute_Contact][View_Weights]

Knowledge Graph Output:

[Entity: Dhanush]

GitHub

dhanush12232002@gmail.com

RAG & Agents

Data Pipelines