Rishabh Dandriyal
Senior AI Engineer
New Delhi,DL
Summary
Senior AI Engineer with 5+ years of experience in data science and applied Generative AI, building production-grade RAG and agentic systems for regulated legal and financial workflows. Strong focus on scalable architectures, reliable LLM orchestration, and data-driven decision automation.
Overview
6
6
years of professional experience
2
2
Certifications
Work History
Senior Consultant - AI Engineer
Ernst & Young
07.2024 - Current
- Created a production-grade legal research assistant (B2B & B2C) using multi-layered Agentic RAG architecture.
- Owned Autonomous Data Indexing pipeline that ingests over ~5 million records in Milvus VectorDB using parent-child chunking, Azure OCR, and Llamaparse.
- Architected large-scale hybrid retrieval on Milvus (TF-IDF + BGE embeddings) with Reciprocal Rank Fusion, improving Top 3 precision from 65% to 92% over a ~5M-document corpus, while maintaining p95 latency of less than 800 ms under production load.
- Productionized a multi-agent LLM workflow using LangGraph to orchestrate planning, retrieval, reasoning, and validation, delivering auditable, citation-backed legal responses with 94% decision accuracy, compliant with Big Four regulatory and audit standards.
- Architected LLM inference and orchestration for production reliability, reducing per-query token usage from approximately 400K to approximately 100K tokens by implementing adaptive chunking, schema-validated outputs (Pydantic), exponential backoff for rate limits, and multi-model fallbacks, resulting in lower latency, stable responses, and predictable costs at scale.
Senior Data Scientist
Nagarro
09.2022 - 07.2024
- Designed and deployed computer vision pipelines for object detection, semantic segmentation, and facial landmark detection using YOLOv7, custom YOLO architectures, and DeepLabv3+ with a ResNet101 backbone, supporting downstream image analysis and automation workflows.
- Implemented OCR pipelines using Azure OCR, combined with regex- and rule-based post-processing, to extract structured text from scanned documents, and developed custom object detection utilities, including Non-Maximum Suppression (NMS) and IoU-based filtering, to improve detection accuracy.
- Improved model performance through data preprocessing (annotation, augmentation, and cleaning), metric-driven evaluation (accuracy, IoU, precision/recall), and hyperparameter tuning to enhance robustness and accuracy.
Data Scientist
Infogain
01.2020 - 09.2022
- Performed data preprocessing, EDA, and feature engineering on large, structured, and text-based datasets to support forecasting, classification, and recommendation systems.
- Built and evaluated machine learning and statistical models, including regression, Random Forest, clustering, and similarity-based NLP models, using TF-IDF and cosine similarity.
- Implemented time-series forecasting solutions and optimized model performance through feature selection, hyperparameter tuning, and metric-driven evaluation.
- Deployed ML models using Streamlit to enable interactive analysis and business-facing demonstrations.
Education
Bachelor of Technology - Computer Science & Engineering
Manav Rachna International Institute of Research And Studies
Faridabad, India 04.2001 -
12 Th - Science
CRPF Public School
New Delhi, India 04.2001 -
10 Th -
CRPF Public School
New Delhi, India 04.2001 -
Skills
Agentic RAG, LangChain, LangGraph, Prompt Engineering , LLM Finetuning,
Certification
Azure fundamentals (AZ- 900)
Timeline
Senior Consultant - AI Engineer
Ernst & Young
07.2024 - Current
Senior Data Scientist
Nagarro
09.2022 - 07.2024
Data Scientist
Infogain
01.2020 - 09.2022
Bachelor of Technology - Computer Science & Engineering
Manav Rachna International Institute of Research And Studies
04.2001 -
12 Th - Science
CRPF Public School
04.2001 -
10 Th -
CRPF Public School
04.2001 -