Looking for a job

Srashti Shetty

@srashtishetty

AI Engineer at SH PROPTECH LIMITED who deployed an AI platform for professional reports and built a RAG spatial planning tool.

India

Message

At SH PROPTECH LIMITED, I designed and deployed an AI platform that automates professional report generation from environmental, heritage, and risk data. I also built a RAG-powered spatial planning tool that uses regulatory documents to generate site-specific strategies.

I developed PDF and GeoJSON ingestion pipelines for a semantic-search knowledge base using Azure AI Search and vector embeddings. I also created an AI evaluation framework and response optimizations for cost and output quality.

At TELENETIX PRIVATE LIMITED, I deployed AI solutions for speech recognition and semantic search, and worked on ML training and inference pipelines. Earlier, as a Data Science Intern at A.N.D INTERNS, I analyzed datasets and developed regression and classification workflows.

Experience

Work history, roles, and key accomplishments

AI Engineer

SH Proptech Limited

Aug 2025 - Mar 2026 (7 months)

Designed and deployed an end-to-end AI platform automating multi-section professional report generation by integrating heterogeneous environmental, heritage, and risk data with LLM reasoning and structured output pipelines. Built a RAG-powered spatial planning tool and developed PDF and GeoJSON ingestion pipelines into a semantic-search knowledge base.

AI Engineer Intern

Telenetix Private Limited

Feb 2025 - May 2025 (3 months)

Deployed AI solutions focused on speech recognition and semantic search, integrating backend systems with ML inference services. Implemented and maintained scalable ML training and inference pipelines, optimising GPU-backed model training for performance and efficiency.

Speech Recognition Semantic Search ML Pipelines GPU Optimization Transformers Fine Tuning Vector Retrieval NLP Python PyTorch TensorFlow

Data Science Intern

A.N.D Interns

Oct 2023 - Dec 2023 (2 months)

Performed EDA on real-world datasets to identify customer behaviour patterns and data-quality issues, informing downstream modelling strategies. Developed regression and classification workflows with feature engineering and hyperparameter tuning, and evaluated models using precision, recall, F1-score, and AUC-ROC.