Looking for a job

ARYAN AGARWAL

@aryanagarwal2

Machine Learning Engineer at 26 Miles Capital, where I built an LLM research assistant that reduced research turnaround time by 60%.

India

Message

What I'm looking for

I'm interested in building reliable, measurable AI systems for real-world use.

At 26 Miles Capital, I architected an internal LLM/RAG research assistant that extracts, summarizes, and scores sentiment from earnings call transcripts and macro filings, reducing research turnaround time by 60%.

I also developed predictive modeling and time-series forecasting pipelines over tick and orderbook data, engineering more than 50 alpha features. My models achieved 40%+ annualized returns with Sharpe > 2.5, and my Python/C++ feature computation and inference pipelines met strict <10ms execution SLAs.

For an Enterprise Agentic RAG Platform project, I built hybrid retrieval, query transformation, and automated evaluation with Ragas. Earlier, as a Machine Learning Intern at Unacademy, I developed Transformer-based NLP intent classification pipelines and built offline evaluation suites and containerized REST APIs.

Experience

Work history, roles, and key accomplishments

Current

Quantitative Researcher / Machine Learning Engineer

Current

26 Miles Capital

Jun 2022 - Present (4 years 5 months)

Architected an internal LLM/RAG research intelligence assistant to extract, summarize, and score sentiment from earnings call transcripts and macro filings, reducing research turnaround time by 60%. Built end-to-end predictive modeling and time-series forecasting pipelines over tick and orderbook data, engineering 50+ alpha features.

Lang Chain ChromaDB XGBoost LightGBM Python C++Redis Time Series Feature Engineering Portfolio Optimization VARs

Machine Learning Intern

Unacademy

Nov 2021 - Jan 2022 (2 months)

Developed NLP intent classification pipelines with Transformer models for customer query routing. Built offline evaluation suites and deployed containerized REST APIs via FastAPI and Docker.

NLP Transformers fastAPI Docker Evaluation