For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

LoginCreate account
Invite to Job
Ardra Rai

Ardra Rai

LLM Systems Analyst | RAG Evaluation, Grounding & Retrieval Optimization

India flagBengaluru, India
$15.00/hrEntry LevelLabel StudioDoccanoScale AI

Key Skills

Software

Label StudioLabel Studio
DoccanoDoccano
Scale AIScale AI
Google Cloud Vertex AIGoogle Cloud Vertex AI
ArgillaArgilla

Top Subject Matter

LLM Systems & RAG Evaluation (Grounding, Retrieval, Hallucination Analysis)
AI Reliability & Research Analysis (Structured Output, Evidence-Based Generation)
Text Data Annotation & NLP Tasks (Classification, Extraction, QA)

Top Data Types

TextText
DocumentDocument
Computer Code ProgrammingComputer Code Programming

Top Task Types

Classification
Entity Ner Classification
Question Answering
Text Summarization
Evaluation Rating
RLHF
Prompt Response Writing SFT
Computer Programming Coding

Freelancer Overview

RAG Behavior Analysis System – LLM Evaluation and System Analysis. Brings 2+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal and Proprietary Tooling. Education includes Bachelor of Engineering, Dr. Ambedkar Institute of Technology (2023). AI-training focus includes data types such as Text and labeling workflows including Evaluation and Rating.

Entry LevelEnglishHindi

Labeling Experience

Research Gap Extraction System – LLM Output Labeling and Evaluation

Text
Engineered a constrained LLM system to extract and label research gaps directly from source documents with clear justification markers. Applied systematic signal-based evaluation and refusal testing to enhance factual reliability and prevent unsupported outputs. Structured system outputs for downstream utility with explicit explanation labeling. • Implemented justification-based evaluation of LLM-generated research gaps. • Applied filtering and refusal mechanism for reliable labeling. • Focused on factual accuracy in output reliability tasks. • Used local LLM inference and Python for system construction.

Engineered a constrained LLM system to extract and label research gaps directly from source documents with clear justification markers. Applied systematic signal-based evaluation and refusal testing to enhance factual reliability and prevent unsupported outputs. Structured system outputs for downstream utility with explicit explanation labeling. • Implemented justification-based evaluation of LLM-generated research gaps. • Applied filtering and refusal mechanism for reliable labeling. • Focused on factual accuracy in output reliability tasks. • Used local LLM inference and Python for system construction.

2026 - Present

RAG Behavior Analysis System – LLM Evaluation and System Analysis

Text
Developed and analyzed a Retrieval-Augmented Generation (RAG) pipeline to systematically evaluate output behavior under varying chunking, retrieval, and context construction strategies. Conducted controlled AI output evaluations using local LLMs to observe response stability and grounding. Added metrics and observability to rate retrieval quality and context accuracy across multiple RAG configurations. • Evaluated retrieval precision and context utility in LLM-based tasks. • Conducted experiments to analyze context construction effects on output. • Used local inference (Ollama) to support LLM evaluation tasks. • Analyzed system bottlenecks by rating retrieval and context structure.

Developed and analyzed a Retrieval-Augmented Generation (RAG) pipeline to systematically evaluate output behavior under varying chunking, retrieval, and context construction strategies. Conducted controlled AI output evaluations using local LLMs to observe response stability and grounding. Added metrics and observability to rate retrieval quality and context accuracy across multiple RAG configurations. • Evaluated retrieval precision and context utility in LLM-based tasks. • Conducted experiments to analyze context construction effects on output. • Used local inference (Ollama) to support LLM evaluation tasks. • Analyzed system bottlenecks by rating retrieval and context structure.

2026 - Present

Education

D

Dr. Ambedkar Institute of Technology

Bachelor of Engineering, Computer Science and Engineering

Bachelor of Engineering
2023

Work History

F

Freelance

Operations & Data Support

Bengaluru
2025 - 2026