Looking for a job

Aaditi Kumari

@aaditikumari

Message

I evaluate LLMs and build reproducible RAG and agentic ML systems.

India

Message

What I'm looking for

I'm looking to build and evaluate ML and LLM systems, contribute to reproducible data and model-evaluation workflows, and work cross-functionally to turn assessment findings into better model decisions.

At Ethara.AI, I evaluated multi-turn LLM outputs using pairwise comparisons and weighted rubrics for instruction-following, factual accuracy, reasoning, and safety.

I completed 900+ AI evaluation tasks across multimodal and LLM benchmarking projects, turning model outputs into clear assessments that informed model selection and improvement.

I curated 200+ SFT training trajectories, completed 700+ structured factual-accuracy annotations, and benchmarked 200+ AI-generated image pairs across five quality dimensions.

I also build end-to-end Python, LangChain, and RAG systems, including hybrid retrieval pipelines with automated evaluation and CI gating, as well as research agents and React-based web applications.

Experience

Work history, roles, and key accomplishments

LLM Post-Training Intern

Ethara.AI

Jan 2026 - Jul 2026 (6 months)

Evaluated multi-turn LLM outputs using pairwise comparison and weighted rubrics, curated SFT training trajectories, and completed 900+ AI evaluation tasks across multimodal and LLM benchmarking projects.