Dhanashri G

Senior AI/ML Engineer with 5+ years of experience architecting and delivering production-grade AI systems that drive measurable business outcomes. Recognized for end-to-end ownership of LLM applications, RAG pipelines, and agentic AI systems from design and infrastructure decisions through deployment and performance optimization at enterprise scale. Proven track record of translating complex AI capabilities into tangible results: reducing operational costs, accelerating response times, and enabling intelligent automation across 10,000+ user interactions monthly.

Combines deep technical expertise in LangChain, Hugging Face, and vector databases with the strategic perspective needed to align AI initiatives with business goals. Now focused on senior and lead AI engineering roles where architecture ownership, cross-functional influence, and production impact define success.

Senior AI Engineer / Automation Developer - EvolutIA

(2025-01)

• Owned end-to-end architecture of a Kore.ai conversational AI platform handling 10,000+ monthly user interactions from intent modeling and NER design through API integration and production deployment improving response efficiency by 35% and reducing average handle time by 25%.
• Drove the technical design of a production RAG pipeline using FAISS vector indexing, sentence embeddings, and LangChain, improving enterprise query accuracy by 40%; made key architectural decisions on chunking strategy, retrieval ranking, and LLM orchestration.
• Led integration of AI services across 5+ enterprise systems via REST APIs, defining the integration architecture and ensuring real-time data consistency across customer-facing applications.
• Established reusable Automation Anywhere A360 component library for data extraction and system integrations, cutting manual processing time by 50% and accelerating delivery of future automation initiatives.
• Partnered with business stakeholders to translate operational requirements into AI-driven solutions, influencing product roadmap decisions and demonstrating ROI through measurable efficiency gains.

Generative AI Engineer - MUFG

(2023-04 - 2024-12)

• Architected a Retrieval-Augmented Generation (RAG) pipeline using LangChain, FAISS, and Chroma to power intelligent search across 1M+ internal financial documents including regulatory filings, compliance policies, and audit reports; reduced document retrieval time by 35%.
• Engineered LLM-powered automation for regulatory compliance workflows, applying advanced prompt engineering to extract structured data from unstructured financial documents across 10+ compliance use cases, reducing manual review effort by 40%.
• Built and deployed a conversational AI assistant for internal banking operations using Hugging Face Transformers and semantic embeddings, enabling natural language querying of transaction data and improving analyst productivity by 30%.
• Developed supervised ML models (Random Forest, Gradient Boosting) for anomaly detection and fraud risk scoring on high-volume financial transaction datasets, improving fraud detection precision by 22% while reducing false positives.
• Containerized and deployed AI microservices on AWS using Docker and CI/CD pipelines, ensuring secure, scalable, and high-availability production environments compliant with financial data governance standards.

Data Analyst II - UnitedHealth Group

(2020-06 - 2022-07)

• Applied supervised ML (Random Forest, Logistic Regression) to predict 30-day patient readmissions, contributing to a 9% reduction presented findings and methodology to clinical and operational stakeholders to drive adoption.
• Designed and optimized SQL/Python ETL pipelines for enterprise healthcare reporting, cutting reporting delays by 20% while maintaining HIPAA compliance across all data flows.
• Built Tableau dashboards for HEDIS and CMS quality metrics, enabling real-time decision-making for senior clinical and operational leaders.
• Identified cost drivers across large-scale claims and clinical datasets, directly supporting cost-reduction initiatives that delivered a 12% decrease in unnecessary expenditure.

Master of Science - Information Systems - Central Michigan University, College of Business Administration (2024)

Bachelor of Technology - Computer Science & Engineering - Siddhartha Institute of Engineering & Technology (2020)