Open to opportunities

Anant Srivastava

@anantsrivastava

Message

I build LLM agents, financial ML pipelines, and production data automation.

India

Message

What I'm looking for

I'm looking to build production AI, ML, and data products where I can own design through deployment, work on complex financial or enterprise problems, and deepen my experience with LLM systems, model evaluation, and scalable APIs.

At Fidelity Investments, I lead Fixed Income workstreams spanning hedge-fund ML modeling and LLM-powered data-quality root-cause analysis. I designed a GenAI pipeline that classifies 7,000+ hedge funds and forecasts next-month returns using 900+ financial and metadata features.

I built an RCA agent that reads failed data-quality jobs, traces Oracle dependencies, retrieves schema context, and maps errors to failing SQL blocks. Its REST API reduced diagnosis time from 4–5 hours to under 10 minutes.

Previously, I automated multi-vendor data evaluation at Fidelity and built DCGAN and Pix2Pix pipelines at DRDO CABS. I also build financial analytics and hybrid RAG systems, from IPO document reasoning to bond-risk dashboards.

Experience

Work history, roles, and key accomplishments

Current

AI Software Engineer

Current

Fidelity Investments

Aug 2025 - Present (1 year)

Led two concurrent Fixed Income workstreams covering hedge-fund ML modeling and LLM-powered Data Quality RCA automation, owning design, implementation, evaluation, and API delivery. Designed a two-stage Gen AI pipeline to classify 7,000+ hedge funds and forecast next-month returns, and built an LLM-powered RCA agent that reduced diagnosis time from 4-5 hours to under 10 minutes.

Python SQL Machine Learning LLM RAG Random Forest Gradient Boosting XGBoost SVM REST APIs Oracle Docker

Data Science Intern

Fidelity Investments

Jan 2025 - Jun 2025 (5 months)

Built an LLM agent for Best-of-Breed data architecture to evaluate multi-vendor datasets, reducing 9 manual analyst steps to one workflow. Containerized and deployed the workflow using Docker and Jenkins CI/CD, reducing processing time by 35% and improving reconciliation accuracy by 12%.

Python LLM Docker CI CD Data Analysis Automation

Research Intern

DRDO CABS

Sep 2024 - Jan 2025 (4 months)

Engineered DCGAN and Pix2Pix cGAN pipelines for ISAR and ship image generation, increasing dataset availability by 25% and reducing manual design effort by 40%. Built an evaluation and visualization pipeline using SSIM, FID, and PSNR to benchmark GAN outputs.

Python Gan Image Processing SSIM PSNR