Parv J.
Mid-Level AI Engineer
Open to offersNew to PlatformAbout
Parv J. is a seasoned AI Engineer with a Bachelor of Technology in Instrumentation Engineering from the Indian Institute of Technology Kharagpur. Specializing in the integration of Artificial Intelligence technologies, Parv spearheaded the development of AI agent evaluation frameworks at VectorShift, enhancing reliability via Python and Rust. His notable achievements include crafting a real-time collaborative editor predicated on sophisticated concurrency controls. Previously, as a Data Science Engineer at Honeywell Technology Solutions, Parv architected a scalable data pipeline optimized for GenAI, significantly cutting processing times for a sizable user base. During his internship at Siemens Technology and Services, he pioneered the use of Langchain for autonomous LLM agents, achieving remarkable accuracy in strategic document processing. Parv also demonstrated his prowess in cloud security with CloudDefense.AI, employing K-modes clustering and Isolation Forest methodologies for precise anomaly detection. His technical proficiency extends across a plethora of tools and languages, including PySpark, PostgreSQL, and Tensorflow, making him a versatile asset in the technology and engineering domains.
Experience
AI Engineer
VectorShift [YCombinator 23] · 2026 — PresentConstructed an AI agent evaluation framework to assess agent performance, focusing on tool accuracy and response correctness. Improved agent reliability through prompt tuning and structured inputs/outputs, and expanded integrations with custom API hooks. Created a real-time collaborative editor that supports human-agent co-editing utilizing patch/line/V4A-based edits and delta generation. Applied concurrency control techniques, including delta computation and Operational Transformation for resolving conflicts and ensuring consistency.Data Science Engineer 2
Honeywell Technology Solutions · 2024 — 2025Designed a scalable data pipeline tailored for a GenAI-enabled AMS portal to optimize processing time. Developed an automated system for incremental data ingestion for over 500,000 Salesforce records, incorporating change data capture and PII encryption. Created a vectorization pipeline utilizing PostgreSQL and pgvector extensions, significantly enhancing the response time for ML model queries. Established a scalable telemetry validation framework capable of processing large amounts of real-time data.Gen AI Intern
Siemens Technology and Services Pvt. Ltd. · 2023 — 2024Engineered a production-ready RAG pipeline using Langchain and Qdrant vectorDB for strategic document retrieval. Developed an autonomous AI agent leveraging the Gemini API, capable of handling over 1000 requests using function-calling and chain-of-thought prompting. Created an automated document processing system that incorporates web scraping, PDF extraction, and a multi-agent RAG system utilizing Langchain. Improved decision-making efficiency with automated reasoning and structured generation of LLM outputs for strategic frameworks.Machine Learning Intern
CloudDefense.AI · 2023 — 2023Developed a hybrid anomaly detection system using K-modes clustering and Isolation Forest, achieving high precision in cloud threat detection. Created a scalable preprocessing pipeline incorporating MinMaxScaler normalization and t-SNE embedding for log feature extraction. Improved validation accuracy through visualizations and cross-validation of anomalies against predetermined rules.
Skills & Expertise
Education
- Bachelor of Technology in Instrumentation EngineeringIndian Institute of Technology Kharagpur · 2020 — 2024
- All India Senior School Certificate ExaminationDelhi Public School, Jaipur · 2018 — 2020