About me
I’m Daivik Pelathur, an AI and Data Engineer based in Toronto, Ontario. I specialise in building production-grade AI systems — from RAG pipelines and LLM orchestration to ETL infrastructure and BI reporting. My work sits at the intersection of data engineering and generative AI: turning messy, unstructured information into reliable, intelligent outputs.
Right now I’m actively developing two projects. Aequitas-Fin is a financial intelligence agent that uses dual-embedding RAG retrieval to analyse dense financial documents — achieving a 0.816 semantic similarity benchmark across 500+ reports. LectureForge is a document-to-video AI pipeline that transforms slide decks and PDFs into narrated educational videos, reducing content inconsistencies by ~40% through slide-aware chunking and LLM summarisation.
Before focusing on AI systems, I worked as a Cloud Data Associate at Accenture, where I built large-scale ETL pipelines on Azure Databricks and supported ML model deployments for enterprise clients. I hold a Graduate Certificate in Artificial Intelligence & Machine Learning from Lambton College (2025) and a BTech in Computer Science Engineering from SRM Institute of Technology (2023). I’ve also contributed to healthcare AI research — co-authoring a study that achieved 94% diagnostic accuracy for cognitive decline detection while reducing assessment burden by 73%, fully compliant with PHIPA standards.
Outside of building things, you’ll find me exploring Toronto’s neighbourhoods, or deep in a rabbit hole about the latest developments in LLM architecture. I believe the best technical work happens when you genuinely understand the problem you’re solving — which means a lot of conversations, a lot of questions, and occasionally getting things wrong before getting them right.
If you’re working on something interesting in data, AI, or both — I’d love to hear about it. Reach me at daivikpelathur.career@gmail.com or connect on LinkedIn.