Notes

LOVISH -
OBJECTIVE -
Enhance system stability, data integrity, and observability through regression validation, automated reconciliation, and structured logging.
Improve multi-state compliance accuracy via output benchmarking, variance analysis, and anomaly detection across pipelines.

Worked on multi-state compliance output benchmarking against expected rule logic with automated variance detection, anomaly flagging, and traceable QA validation workflows.
Worked on enhancing pipeline observability and auditability by implementing structured logging, decision-boundary tracing, and failure-path instrumentation across RetrievalQA and compliance processing layers.
Working on Louisiana corpus ingestion (Building, Plumbing, Architecture, Residential, Renovation) with preprocessing, embedding generation, and ChromaDB vector store integration for rule retrieval.
Working on integration of North Carolina Structural and Architectural law corpora and cheat-sheet–linked retrieval mappings into the compliance engine.


KALASH -
objective -
Build a stable, backward-compatible compliance pipeline with multi-state regression checks, variance detection, and standardized rule normalization.
Enable scalable compliance retrieval by ingesting Louisiana law data into ChromaDB with structured preprocessing and cheat-sheet–linked RetrievalQA.

Working on designing and implementing scraping and preprocessing pipelines for Louisiana law documents, including structured text extraction, semantic chunking, and metadata enrichment for downstream vector ingestion.
Working on engineering ChromaDB-backed vector stores for law corpora with cheat-sheet linkage to enable scalable, traceable RetrievalQA-driven compliance evaluation.
Working on merging Connecticut 2022 amendments with the 2021 IBC baseline and constructing the corresponding normalized vector database for unified rule retrieval.
Working on integrating Connecticut Architecture and Structural law corpora with aligned cheat sheets into the compliance pipeline, alongside Alaska multi-domain law scraping (Architecture, Mechanical, Renovation, Fuel Gas).


Worked Completed Louisiana file scraping for Building, Plumbing, Architecture, Residential, and Renovation and integrated the data into the pipeline.

Worked on integrated North Carolina Structure and Architecture law files and cheat sheets provided by Scott into the compliance workflow.

worked Merged Connecticut 2022 amendments with the 2021 IBC and successfully built the corresponding vector database.

working on Connecticut Architecture and Structural law file and cheat sheet integration along with Alaska file scraping for Architecture, Mechanical, Renovation, and Fuel Gas.