Piyush Gupta

Data Scientist and AI Engineer with 2+ years of production experience building AI-powered products including LLM based tools using ChatGPT (GPT-4.1, GPT-5), Azure AI Foundry, and open-source models (Mistral, Phi-3) deployed at enterprise scale in a global engineering firm.

Shipped real features: a text summarization pipeline, a sequence-aware recommendation engine, and chart-to-text AI tools using vision models. Published 3 peer-reviewed research papers in AI/ML. Combines engineering rigour with measurable business impact from a $3 billion client proposal to a live international deployment in Amsterdam.

Data Scientist at Fluor Daniel India Private Limited (2026-03 – Present)

Selected for a company-wide AI initiative; deployed to Fluor B.V. Amsterdam for 6 weeks to build production grade LLM-based tools and recommendation engines.

• Selected for a company-wide AI initiative; deployed to Fluor B.V. Amsterdam for 6 weeks to build a production grade LLM-based text summarization feature for the firm's proprietary platform.
• Designed a sequence-aware recommendation engine using LSTM (TensorFlow) to personalise user navigation across 200+ data visualisation assets.
• Built two AI proof-of-concept tools for chart-to-text conversion: one using Microsoft Phi-3 Vision (open-source vision LLM via Hugging Face / ONNX) and one using GPT-4.1 and GPT-5 via Azure AI Foundry, generating readable summaries from chart images.

Associate Data Scientist at Fluor Daniel India Private Limited (2024-07 – 2026-03)

Built interactive dashboards, Streamlit apps, and automated data pipelines for real-time project health diagnostics and risk forecasting.

• Built interactive data dashboards using Plotly.js and D3.js, embedded into the company's proprietary platform for real-time project health diagnostics.
• Developed two Streamlit apps to automate data ingestion and cleansing from Coreworx and Contracts systems into SQL production databases.
• Built a Control System Tower demo in RShiny with real-time risk forecasting visuals used directly in a $3 billion client proposal.
• Automated R scripts using ggplot2 and Plotly to benchmark live projects against more than 300 historical datasets, enabling trend analysis and risk flagging.
• Collaborated with Data Analysts on deep-dive project health assessments and risk reporting for senior stakeholders

Data Scientist (Intern) at Fluor Daniel India Private Limited (2024-03 – 2024-07)

Built ETL pipelines and automated time-series data preparation for project data extraction and validation.

• Built ETL pipelines in Python (Pandas) and SQL to extract, clean, and validate project data from Coreworx and Contracts systems.
• Automated time-series data preparation for Primavera P6 schedule data in R, reducing manual processing time by 30%.

Software Developer (Intern) at CRIS (2023-07 – 2023-09)

Built real-time tracking system and optimized SQL query pipelines for railway asset management.

• Built real-time tracking system using Flutter, improving operational visibility for railway assets.
• Optimized high-volume SQL query pipelines, reducing data retrieval latency and improving system performance.

Bachelor of Technology in Computer Science – Guru Gobind Singh Indraprastha University (2020-08 – 2024-07)