Sankalp Sinha profile photo 1 on Tal

Sankalp Sinha

Data Scientist||Temenos||CMI

Bengaluru

6+ years of work experience

SimplifyX
Mar 20251 yr 6 mosPresent

Bengaluru

Regulatory Intelligence Platform: architected an end-to-end system automating RIM data collection and scrutiny across Medicaid, CMS, Federal Register, Transmittals, and 20+ regulatory sources using LLM-powered web agents and structured extraction pipelines eliminating 24 hrs/month of manual research (≈300 hrs/year saved). Document Agent: designed and built the backend for a Document Agent on the in-house LLM orchestration platform that extracts precise field values from unstructured PDFs (claims forms, policy docs, transmittals) given a field name + description. RAG Pipelines: implemented and productionized RAG pipelines integrating LLMs with vector databases, achieving 85%+ reduction in hallucination and 4× faster response times on regulatory queries. Agentic Systems & Observability: built agentic workflows using LangChain and LangGraph; deployed agents with Aegra and instrumented pipelines end-to-end with OpenTelemetry (OTel) for tracing, monitoring, and observability of LLM/GenAI applications. Cross-functional delivery: collaborated with the Platform team to build automated multi-LLM workflows for document understanding and data extraction.
Temenos India Pvt. Ltd. — AI Business Unit
Nov 20204 yrs 4 mosFeb 2025

Bangalore Urban

Pioneered RAG‑based autonomous SQL query generator: LLMs dynamically create complex SQL with natural‑language explanations; adopted across multiple banking clients and became a core differentiator in negotiations. Built production RAG pipeline integrating a proprietary vector database with in-house GPU-hosted models, delivering enhanced retrieval accuracy for financial‑domain queries. Developed a custom Named Entity Recognition engine using fine‑tuned transformer models on scarce, domain‑specific banking/regulatory documents; outperformed off‑the‑shelf SpaCy/BERT models by 18 F1 points. Created a product‑recommendation engine with an in‑house explainable‑AI layer; deployed via FastAPI + Docker + KServe and actively supported pre‑sales efforts, helping the team secure high‑value, profitable contracts with major banks. Led end‑to‑end NLP payment‑message flagging system for a Tier‑1 European bank using FastAPI microservices and Docker; the solution was a decisive factor in winning the client contract.
Temenos
May 20193 mosJul 2019

Bengaluru Area, India

I got the opportunity to work on fraud detection in payment transactions, Explainable AI and building model for NER, with major focus on most critical aspect of Enterprise class AI -XAI (Explainability). Further I also got to know the ABC's of RESTFUL API development. Awarded PPO
Temenos India Pvt. Ltd. — AI Business Unit
May 20193 mosJul 2019

Engineered XAI layers using SHAP and LIME to interpret Fraud Detection and NER models, ensuring enterprise-grade model transparency, trust, and regulatory compliance.

Sankalp’s AI Stack

claude
claude
cursor
cursor
langchain
langchain
langgraph
langgraph
n8n
n8n
huggingface
huggingface

Education

Chennai Mathematical Institute

Master of Science, Data Science

Chennai Mathematical Institute

20182 yrs2020
BIET Jhansi

Bachelor of Technology, Computer Engineering

BIET Jhansi

Join Sankalp and 10k+ other professionals on

Tal Boss

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

Download Tal Boss

Sankalp Sinha on Tal

Headline

Data Scientist||Temenos||CMI

Location

Bengaluru

AI stack

claude

cursor

langchain

langgraph

n8n

huggingface

Work experience

Senior Data Scientist at SimplifyX, Mar 2025 - Present · 1 yr 6 mos, Bengaluru, Regulatory Intelligence Platform: architected an end-to-end system automating RIM data collection and scrutiny across Medicaid, CMS, Federal Register, Transmittals, and 20+ regulatory sources using LLM-powered web agents and structured extraction pipelines eliminating 24 hrs/month of manual research (≈300 hrs/year saved). Document Agent: designed and built the backend for a Document Agent on the in-house LLM orchestration platform that extracts precise field values from unstructured PDFs (claims forms, policy docs, transmittals) given a field name + description. RAG Pipelines: implemented and productionized RAG pipelines integrating LLMs with vector databases, achieving 85%+ reduction in hallucination and 4× faster response times on regulatory queries. Agentic Systems & Observability: built agentic workflows using LangChain and LangGraph; deployed agents with Aegra and instrumented pipelines end-to-end with OpenTelemetry (OTel) for tracing, monitoring, and observability of LLM/GenAI applications. Cross-functional delivery: collaborated with the Platform team to build automated multi-LLM workflows for document understanding and data extraction.

Data Scientist at Temenos India Pvt. Ltd. — AI Business Unit, Nov 2020 - Feb 2025 · 4 yrs 4 mos, Bangalore Urban, Pioneered RAG‑based autonomous SQL query generator: LLMs dynamically create complex SQL with natural‑language explanations; adopted across multiple banking clients and became a core differentiator in negotiations. Built production RAG pipeline integrating a proprietary vector database with in-house GPU-hosted models, delivering enhanced retrieval accuracy for financial‑domain queries. Developed a custom Named Entity Recognition engine using fine‑tuned transformer models on scarce, domain‑specific banking/regulatory documents; outperformed off‑the‑shelf SpaCy/BERT models by 18 F1 points. Created a product‑recommendation engine with an in‑house explainable‑AI layer; deployed via FastAPI + Docker + KServe and actively supported pre‑sales efforts, helping the team secure high‑value, profitable contracts with major banks. Led end‑to‑end NLP payment‑message flagging system for a Tier‑1 European bank using FastAPI microservices and Docker; the solution was a decisive factor in winning the client contract.

AI Intern at Temenos, May 2019 - Jul 2019 · 3 mos, Bengaluru Area, India, I got the opportunity to work on fraud detection in payment transactions, Explainable AI and building model for NER, with major focus on most critical aspect of Enterprise class AI -XAI (Explainability). Further I also got to know the ABC's of RESTFUL API development. Awarded PPO

Data Science Intern at Temenos India Pvt. Ltd. — AI Business Unit, May 2019 - Jul 2019 · 3 mos, Engineered XAI layers using SHAP and LIME to interpret Fraud Detection and NER models, ensuring enterprise-grade model transparency, trust, and regulatory compliance.

Education

Master of Science, Data Science, Chennai Mathematical Institute

Bachelor of Technology, Computer Engineering, BIET Jhansi