Open to opportunities

Nature Trolly

@naturetrolly

Message

I build production-ready LLM, RAG, and federated learning systems for accurate, privacy-preserving AI.

India

Message

What I'm looking for

I'm looking to build reliable AI products involving LLMs, RAG, prompt engineering, and privacy-preserving machine learning, while collaborating with engineering teams on scalable production systems.

I've built and optimized LLM prompt pipelines at Scalar AI, improving response accuracy and contextual relevance by 28%.

I developed few-shot prompting frameworks and instruction-tuning datasets to reduce hallucinations in domain-specific technical queries. I also fine-tuned Llama-3 and Mistral with LoRA and QLoRA on custom domain corpora.

I collaborated with engineering teams to integrate LLM endpoints into REST microservices handling more than 50,000 daily inference requests with minimal latency.

My projects include an auditable privacy-preserving federated learning framework, a hybrid-search RAG system with 35% higher top-3 retrieval relevance, and an interactive dashboard for automated data insights and predictive visualization.

Experience

Work history, roles, and key accomplishments

Machine Learning & Prompt Engineering Intern

Scalar AI

Jun 2025 - Dec 2025 (6 months)

Engineered and optimized prompt pipelines for LLMs, improving response accuracy by 28%. Developed few-shot frameworks and fine-tuned open-source models using LoRA/QLoRA, integrating endpoints into microservices handling 50,000 daily requests.

Prompt Engineering LLM Fine tuning LoRAs QLoRA Mistral REST APIs Microservices Python