Aryan Sangam profile photo 1 on Tal

Aryan Sangam

Data Scientist@Accordion | GenAI • LLMs • RAG • Agentic AI • Azure OpenAI • Machine Learning • Python • SQL

Hyderabad

2+ years of work experience

Accordion
Jan 20268 mosPresent

• Built and deployed a multi-agent AI Store Intelligence Platform on Azure using LangGraph, FastAPI, Azure OpenAI, Azure SQL, Azure AI Search (RAG), Azure Functions, and Docker, enabling natural language analytics over structured sales data, vision-derived operational metrics, and customer reviews. • Built a proof-of-concept AI-powered data extraction pipeline by recursively crawling Chipotle’s online store locator, constructing a hierarchical graph (Country → State → Store → Menu Links), and integrating Azure OpenAI APIs and LangChain agents to automate storewise menu pricing extraction into structured Excel reports. • Engineered a Snowpark-based data ingestion pipeline to replace Fivetran connectors for Salesforce Marketing Cloud, implementing incremental loads with watermarking across 6 event types, reducing ingestion costs by 65% while maintaining 99.8% data accuracy.
Accordion
Jul 20241 yr 6 mosDec 2025

Hyderabad, Telangana, India

• Developed a logistic regression-based SLA prediction model that enabled a Fortune 500 commercial real estate client to proactively identify at-risk work orders, achieving 80% accuracy with an 84% True Negative Rate and improving SLA compliance. • Built and deployed a Python + Gurobi optimization model (Excel-integrated) to standardize pricing for 300+ clients, designing five multipliers and solving for minimal variance from historical spend while automating client allocation across the new structure. • Developed a vendor replacement model for CBRE by analyzing external vendor performance, identifying costly or underperforming vendors, and recommending optimal vendor replacements based on similarities in city operations, work order problem codes, and cost efficiency.
Piramal Capital & Housing Finance Limited
Jul 20236 mosDec 2023

Bengaluru, Karnataka, India

Collaborated with cross-functional teams to prepare data for dashboards using SQL, Python and Spark

Aryan’s AI Stack

codex
codex
cursor
cursor

Education

Birla Institute of Technology and Science, Pilani

Bachelor of Engineering - BE, Electronics and Instrumentation

Birla Institute of Technology and Science, Pilani

20204 yrs2024
Birla Institute of Technology and Science, Pilani

Minor, Data Science

Birla Institute of Technology and Science, Pilani

20222 yrs2024

Join Aryan and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Aryan Sangam on Tal

Headline

Data Scientist@Accordion | GenAI • LLMs • RAG • Agentic AI • Azure OpenAI • Machine Learning • Python • SQL

About

Data Scientist with experience building AI and machine learning solutions across Generative AI, RAG systems, agentic workflows, predictive modeling, optimization, and large-scale data engineering. At Accordion, I have developed AI-powered assistants, automated data extraction pipelines, LLM applications, and predictive analytics models for Fortune 500 clients.

Location

Hyderabad

AI stack

codex

cursor

Work experience

Associate 1 - Data Science at Accordion, Jan 2026 - Present · 8 mos, • Built and deployed a multi-agent AI Store Intelligence Platform on Azure using LangGraph, FastAPI, Azure OpenAI, Azure SQL, Azure AI Search (RAG), Azure Functions, and Docker, enabling natural language analytics over structured sales data, vision-derived operational metrics, and customer reviews. • Built a proof-of-concept AI-powered data extraction pipeline by recursively crawling Chipotle’s online store locator, constructing a hierarchical graph (Country → State → Store → Menu Links), and integrating Azure OpenAI APIs and LangChain agents to automate storewise menu pricing extraction into structured Excel reports. • Engineered a Snowpark-based data ingestion pipeline to replace Fivetran connectors for Salesforce Marketing Cloud, implementing incremental loads with watermarking across 6 event types, reducing ingestion costs by 65% while maintaining 99.8% data accuracy.

Analyst 2 - Data Science at Accordion, Jul 2024 - Dec 2025 · 1 yr 6 mos, Hyderabad, Telangana, India, • Developed a logistic regression-based SLA prediction model that enabled a Fortune 500 commercial real estate client to proactively identify at-risk work orders, achieving 80% accuracy with an 84% True Negative Rate and improving SLA compliance. • Built and deployed a Python + Gurobi optimization model (Excel-integrated) to standardize pricing for 300+ clients, designing five multipliers and solving for minimal variance from historical spend while automating client allocation across the new structure. • Developed a vendor replacement model for CBRE by analyzing external vendor performance, identifying costly or underperforming vendors, and recommending optimal vendor replacements based on similarities in city operations, work order problem codes, and cost efficiency.

Data Engineer at Piramal Capital & Housing Finance Limited, Jul 2023 - Dec 2023 · 6 mos, Bengaluru, Karnataka, India, Collaborated with cross-functional teams to prepare data for dashboards using SQL, Python and Spark

Education

Bachelor of Engineering - BE, Electronics and Instrumentation, Birla Institute of Technology and Science, Pilani

Minor, Data Science, Birla Institute of Technology and Science, Pilani