Rohit Halder

Data Scientist

Bengaluru, Karnataka, India4 yrs 11 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Over 5 years of experience in data-driven solutions.
Expertise in machine learning and natural language processing.
Proven track record of reducing SLA breaches by 80%.

Stackforce AI infers this person is a Data Scientist specializing in Fintech and Banking analytics.

Contact

Skills

Core Skills

Machine LearningData ScienceArtificial Intelligence (ai)

Other Skills

AWS SageMakerAdaptive Forecasting FrameworkAgentic AiAlgorithmsAnalytical SkillsAnalyticsAnomaly DetectionAutoARIMAAutoML-ERPBayesian RegressionBusiness Analytics for CollectionsBusiness InsightsBusiness-to-Business (B2B)Cash Flow ForecastingClient feedback

About

Accomplished Data Scientist with >5 years of expertise in delivering end-to-end data-driven solutions for Fortune 100 clients in Fintech and Banking. Advanced proficiency in Python, statistical modeling, machine learning, deep learning, computer vision, natural language processing and Generative AI with applications in retrieval augmented generation (RAG), specializing in Workforce Forecasting, Account Receivables and Treasury Analytics. Demonstrated ability to drive complex data projects from proof of concept to deployment and productization, with comprehensive skills in data visualization and client-facing technical engagements.

Experience

4 yrs 11 mos

Total Experience

1 yr 10 mos

Average Tenure

1 yr 2 mos

Current Experience

Anz

Data Scientist

Apr 2025 – Present · 1 yr 2 mos · Bengaluru, Karnataka, India · On-site

● Developed and enhanced an Adaptive Forecasting Framework capable of autonomously selecting the best-fit algorithm from FbProphet, Exponential Smoothing, AutoARIMA, TBATS, and heuristic ensemble models, supporting adaptive model switching based on data length and seasonality. The framework manages a portfolio of nearly 80 live forecasting models, each iterating in approximately <5 minutes, enabling rapid experimentation and near real-time updates.
● Designed and implemented a holistic Workforce Forecasting and Backlog Simulation Framework for Customer Resolutions that unified heterogeneous data sources, complaint-level ticket details, global employee information, and stakeholder mappings into a single analytical pipeline. Built a modular architecture consisting of (1) prediction of new incoming tickets, (2) forecasting of expected backlog using Bayesian Regression, Linear Regression, and Time-Series models, (3) derivation of resolution-rate distributions grounded in workshare concepts to represent multi-employee ticket handling, (4) simulation of backlog evolution under varying FTE supply scenarios, and (5) formulation of backlog volume simulation as a loss function to identify the optimal staffing level that minimizes SLA breaches and supports proactive workforce planning, achieving an 80% reduction in SLA breaches and establishing a standardized analytical foundation for resource allocation.

Adaptive Forecasting FrameworkFbProphetExponential SmoothingAutoARIMATBATSBayesian Regression+4

Highradius

3 roles

Software Engineer - II (Data Science)

Promoted

Jan 2024 – Jan 2025 · 1 yr

● Developed a modular Proof of Concept for a GenAI-based Agentic AI (RAG) workflow to automate parts of invoice dispute resolution. The system generated compliance emails upon dispute creation, extracted and interpreted invoice details using natural language processing, and interfaced with the MCP server to retrieve corresponding line-item and dispute data. Implemented logic to validate claim legitimacy by cross-verifying invoice-level details and promotion mappings, demonstrating the feasibility of an AI-assisted collections and dispute management process.
● Engineered a transformative collections optimization module by collaborating cross-functionally with Product Management, UI, and Platform teams, leveraging deep stakeholder insights from collector interviews and client feedback. Successfully developed a data-driven solution that enabled collectors to exceed SLA targets by 10% (CPG industry), improving operational efficiency by 10%
● Led a 6-member team supporting sales initiatives, providing critical L2 technical support, developing advanced analytical frameworks, and delivering compelling POCs that ensured seamless client deployments and accelerated product value demonstration for client acquisition.

GenAI-based Agentic AINLPCompliance emailsInvoice dispute resolutionCollections optimizationArtificial Intelligence (AI)+1

Software Engineer - I

Jul 2023 – Dec 2023 · 5 mos

● Architectured incremental training in Linear Regression using clustered samples, achieving a ~15% accuracy improvement. Developed a Payment Behavior Classification metric that enhanced prediction accuracy by ~12% through month-boundary-aware predictive modeling.
● Developed a scalable on-demand deployment architecture to optimize cost-efficient VM/EC2 instance provisioning, managing 3,700+ weekly prediction jobs, including tasks processing over 10M records.

Linear RegressionPayment Behavior ClassificationPredictive modelingMachine Learning

Associate Software Engineer - II

Jan 2023 – Aug 2023 · 7 mos

● Designed an "AutoML-ERP" endpoint that dynamically assesses ERP data quality and tests feasible forecasting algorithms, reducing modeling timelines by 75% with improved accuracy using Machine Learning Algorithms - Linear Regression, Bagging Linear Regression, Decision Tree Regressor, Random Forest Regressors, Gradient Boost Regressors.
● Conducted hypothesis tests for Granger Causality using Trade Promotion Management data, and developed frameworks for pivotal regressors leading to a 17% improvement in medium-term forecast accuracy.

AutoML-ERPMachine Learning AlgorithmsHypothesis testingMachine Learning