Research & Publications

Advancing the frontiers of AI reasoning, financial ML, and domain-specific language models

Competition Recognition

Award-winning contributions to the AI research community

First Place Winner

Reasoning Dataset Creation Competition 2025

Bespoke Labs, HuggingFace & Together.AI

Recognized for creating the PersonalFinance_v2 dataset, featuring high-quality financial reasoning with explicit chain-of-thought methodologies. The dataset demonstrates advanced reasoning capabilities in personal finance scenarios.

7.04k high-quality reasoning examples
8 diverse financial categories
Explicit chain-of-thought explanations
Dataset Impact
23
Community Likes
Apache-2.0
Open License

Current Research

Active research projects and ongoing investigations

Financial Services Innovation Lab

Georgia Institute of Technology

Active Research

Research Focus

Leading research on generalized financial reasoning models with explicit chain-of-thought methodologies. Focus on US finance corpora curation and ablation studies across different model scales.

Key Contributions

Curating high-quality financial QA pairs with domain expertise
Developing scalable reasoning frameworks for financial applications
Running comprehensive ablation studies across dataset complexity

Upcoming Publications

Research papers under review and preparation

FinNLP 2025 Workshop (EMNLP)

Paper under review

Research on financial reasoning models and document processing optimization, focusing on production-ready implementations for enterprise applications.

Dataset Creation Methodologies

In preparation

Comprehensive methodology paper based on the award-winning PersonalFinance_v2 dataset creation process and competition insights.

Research Interests

Core areas of investigation and expertise

Financial AI & FinTech

Behavior-aware personal finance models
Financial document processing
Enterprise fintech applications

Reasoning Systems

Chain-of-thought methodologies
Dataset creation frameworks
Model training optimization

Production ML

Single-GPU deployment optimization
Scalable AI system architecture
Enterprise model deployment

Academic Research

Undergraduate research and early contributions

Impact of Social Engagement on Crowdfunding Campaigns

IIT Kharagpur • Undergraduate Research Project

August 2022 - April 2023 • 9 months

Research Objectives

Comprehensive analysis of social media influence on crowdfunding campaign success, conducted under the guidance of Dr. Swagato Chatterjee at COEAI, IITKGP.

Methodology

Step 1 Identifying characteristics that inspire social media participation
Step 2 Analyzing relationship between engagement and funding success
Step 3 Connecting social factors to campaign funding outcomes

Key Contributions

Statistical correlation analysis framework
Social media engagement metrics
Crowdfunding success prediction model
Research Repository

Research Impact

1st Place
Competition Winner
Reasoning Dataset 2025
Active
Research Affiliate
Georgia Tech
23
Community Likes
Dataset Recognition

Research Collaboration

Interested in collaborating on financial AI, reasoning systems, or dataset creation? Let's explore research opportunities together.