NK
Niraj Kumar

Niraj Kumar

AI Engineer•ML Engineer•Data Scientist•Risk Consultant

Architecting autonomous agentic AI systems and quantitative risk intelligence. Focused on explainable machine learning, predictive analytics, and high-stakes financial decision engines.

Explore Work
Download CV
GitHub•LinkedIn•
India
• • •
• • •

Interactive 3D Portfolio Guide Agent

A conversational 3D intelligence agent representing Niraj Kumar. Explore his Agentic AI systems, Risk Consulting methodologies, Machine Learning architectures, or background in real time.

Direct Answer (Verified CV Knowledge)
Multi-Agent Credit Risk & Financial Intelligence Suite (Oct '26) [Verified CV Project]: • System Architecture: Architected an AI-powered credit risk assessment system using Python and LangGraph, coordinating five specialized agents to automate borrower verification, fraud detection, and loan evaluation. • Predictive Modeling: Developed LightGBM-based predictive models with SHAP explainability to estimate default probability, calculate expected credit loss (ECL), and identify key credit risk factors. • Regulatory Compliance: Integrated RAG-based regulatory analysis, fairness auditing, and CCAR stress testing to detect lending bias, assess financial risks, and generate transparent credit recommendations. • Tech Stack: Python, LangGraph, LangChain, LightGBM, SHAP, RAG, Vector Database, Pandas, NumPy, Scikit-learn, Basel III/IV, CCAR.
• • •
• • •

About Niraj Kumar

Engineering autonomous multi-agent intelligence, mathematical machine learning models, and quantitative risk consulting architectures.

Niraj Kumar

I am a Computer Science Engineer specialized in Agentic AI architectures, Machine Learning engineering, quantitative Risk Consulting, and advanced Data Science & Analytics.

My engineering portfolio features the Multi-Agent Credit Risk & Financial Intelligence Suite, coordinating five specialized LangGraph agents to automate borrower verification, fraud detection, and regulatory CCAR stress testing with SHAP mathematical explainability.

I also engineered Budget Buddy, an autonomous savings agent built with Google Gemini API using autonomous tool chaining and Plan-Act-Observe workflows. Grounded by 200+ LeetCode DSA problems in Java, I approach complex machine learning and data challenges with deep computational rigor.

7.4 CGPA
B.Tech CSE (LPU)
Specialization in AI & Data Science
200+
DSA in Java
Optimized Algorithmic Solutions
5+
AI & ML Frameworks
LangGraph, Gemini, LightGBM, SHAP
7 Certs
Specializations
Coursera, Udemy & Simplilearn
• • •
• • •

Skills & Technology Stack

Specialized domain expertise spanning autonomous agentic systems, explainable machine learning, risk analytics, and quantitative engineering.

Agentic AI & GenAI Technologies

Autonomous multi-agent orchestration, LLM architectures & RAG

Agentic AILangGraph Multi-AgentLangChainRetrieval-Augmented Generation (RAG)Google Gemini APIModel Context Protocol (MCP)LangSmith TracingAutonomous Tool ChainingPlan-Act-Observe WorkflowsPrompt Architecture
10 competenciesCV Verified →

Machine Learning & Risk Consulting

Supervised & unsupervised learning, explainable AI, stress-testing

LightGBM & XGBoostSHAP ExplainabilityCredit Risk ModelingBasel III / IV ComplianceCCAR Stress TestingScikit-learnExpected Credit Loss (ECL)Fairness & Bias AuditingFeature Engineering
9 competenciesCV Verified →

Data Science, EDA & Analytics

High-dimensional data processing, statistical modeling & BI

PythonPandas & PolarsNumPyExploratory Data Analysis (EDA)Statistical Hypothesis TestingPower BI Executive DashboardsSQL Query OptimizationData Cleaning & SanitationTime Series Analysis
9 competenciesCV Verified →

Core CS & Algorithmic Foundations

Algorithmic problem-solving & rigorous software engineering

Java (200+ LeetCode DSA)Dynamic ProgrammingGraph Algorithms (BFS/DFS)Trees & Binary SearchPythonC / C++SQLData Structures & Algorithms
8 competenciesCV Verified →

Platforms, Databases & Tools

ML engineering pipelines, databases & cloud tooling

GitHub & GitMySQL WorkbenchVector DB (Chroma/FAISS)Power BIJupyter NotebookVS CodeKaggleExcel Advanced Modeling
8 competenciesCV Verified →

Professional & Strategic Leadership

Analytical decision-making and cross-functional leadership

Quantitative Problem-SolvingRisk ManagementStrategic CommunicationCross-Functional CollaborationAdaptabilityLeadership
6 competenciesCV Verified →
• • •
• • •

Projects & Architectures

Showcasing autonomous multi-agent systems, explainable machine learning pipelines, and quantitative financial risk intelligence suites.

Agentic AI & RiskOct '26

Multi-Agent Credit Risk & Financial Intelligence Suite

An enterprise-grade autonomous AI credit risk ecosystem coordinating five specialized agents to automate borrower verification, fraud detection, and credit scoring with regulatory compliance.

Architected an AI-powered credit risk assessment system using Python and LangGraph, coordinating five specialized agents to automate borrower verification, fraud detection, and loan evaluation.
Developed LightGBM-based predictive models with SHAP explainability to estimate default probability, calculate expected credit loss (ECL), and isolate key risk factors.
Integrated RAG-based regulatory compliance analysis, fairness auditing, and CCAR stress testing to detect lending bias, assess financial volatility, and output explainable audit trails.
LangGraphLangChainPythonLightGBMSHAPRAGVector DBPandasScikit-learnBasel III/IVCCAR
Code Repository
Agentic AI & RiskAug '26

Budget Buddy – Autonomous Gemini AI Savings Agent

Autonomous financial savings agent powered by Google Gemini API orchestrating a Plan-Act-Observe workflow with intelligent tool chaining for personalized financial recovery.

Engineered an autonomous AI-powered financial savings agent using Google Gemini API, enabling users to set goals, log deposits, monitor progress, and receive personalized catch-up plans.
Orchestrated a Plan-Act-Observe workflow with autonomous tool chaining, allowing the agent to automatically evaluate savings progress and generate accelerated 3-month recovery strategies when users fall behind schedule.
Automated intelligent savings calculations by tracking deadlines, expected progress, shortfalls, remaining balances, and revised monthly targets through persistent transaction history.
Google Gemini APIPythonFlaskPlan-Act-ObserveJSONAutonomous Agents
Code Repository
Intelligent SystemsNov '26

Interactive 3D Portfolio Guide Agent

Conversational agent architecture featuring real-time 3D WebGL avatar rendering and contextual NLP query resolution to guide recruiters through AI, ML, Data Science, and Risk consulting capabilities.

Engineered an interactive 3D robot avatar using Three.js with procedural floating physics, cursor-tracking head tilt, glowing eye visor, and audio-reactive status pulses.
Integrated an intelligent portfolio guide conversational agent answering queries regarding Agentic AI, Machine Learning, Risk Consulting, and data modeling.
Built modular Node.js API endpoints for low-latency query resolution and real-time knowledge matching.
Three.jsConversational AITypeScriptNode.js APINLP Query Matching
Code Repository
Customer Churn Prediction & Retention Analytics
Machine LearningMay '25

Customer Churn Prediction & Retention Analytics

Machine learning analytics pipeline to predict customer churn in telecommunications and banking, identifying churn drivers and optimizing retention strategies.

Engineered predictive machine learning pipelines comparing Random Forests, Gradient Boosting, and Logistic Regression with 86%+ precision.
Conducted extensive Exploratory Data Analysis (EDA) on customer tenure, billing models, and usage metrics to pinpoint high-risk segments.
Provided actionable recommendations that reduce projected churn and maximize customer lifetime value (LTV).
PythonScikit-LearnPandasNumPyMatplotlibSeaborn
Code Repository
Cryptocurrency Market Volatility & Predictive Insights
Data Science & AnalyticsFeb '25

Cryptocurrency Market Volatility & Predictive Insights

High-dimensional financial analytics framework studying cross-token price movements, liquidity patterns, and macroeconomic correlations across crypto markets.

Analyzed multi-year historical time series across Bitcoin, Ethereum, and major altcoins using rolling statistics and Sharpe ratio evaluations.
Built automated data cleaning and anomaly detection pipelines in Python and Pandas handling volatile market anomalies.
Created dynamic visual dashboards showing volume shifts, volatility clusters, and moving average cross-signals.
PythonPandasMatplotlibStatistical ModelingFinancial Analytics
Code Repository
Healthcare Diagnostics & Clinical Risk Intelligence
Data Science & AnalyticsJan '25

Healthcare Diagnostics & Clinical Risk Intelligence

Epidemiological and diagnostic analysis identifying early biomarkers and health risk indicators to assist clinical decision-making.

Conducted statistical hypothesis testing and feature correlation to isolate critical clinical variables across patient records.
Built visualization models illuminating health trends and patient risk clusters to support early clinical intervention.
Standardized heterogeneous medical datasets with rigorous missing data imputation and outlier sanitation.
PythonPandasSeabornStatisticsHypothesis Testing
Code Repository
• • •
• • •

Education & Academic Training

Formative foundation in Computer Science, algorithmic problem solving, and specialized data analytics.

Since Aug '23CGPA: 7.4

Bachelor of Technology - Computer Science and Engineering

Lovely Professional University (LPU)
Punjab, India

• Core focus on Artificial Intelligence, Machine Learning, Data Science, and Algorithmic Problem Solving.

• Active contributor to AI/ML research groups, coding communities, and open-source initiatives.

Verified CredentialAcademic Record
Apr '19 - Mar '2182.6%

Senior Secondary (Intermediate)

Chattradhari High School
Dalsinghsarai, Bihar

• Rigorous foundation in Mathematics, Physics, and Analytical Sciences.

Verified CredentialAcademic Record
Apr '18 - Mar '1980.8%

Secondary School (Matriculate)

Chattradhari High School
Dalsinghsarai, Bihar

• Distinction in Mathematics and Science.

Verified CredentialAcademic Record

Core Data Analytics & Business Intelligence Training

Lovely Professional University (LPU) • Punjab, India (Jun '25 - Jul '25)

Learned core data analytics techniques by collecting, cleaning, and processing datasets using Python, Excel, and MySQL to build a strong foundation in data handling and interpretation.
Created interactive visual dashboards in Power BI and applied analytical methods to interpret trends, summarize insights, and support data-driven decision-making.
Mastered: Power BI, MySQL, Excel, Python, Exploratory Data Analysis, Visualizations, and Executive Dashboards.
• • •
• • •

Professional Certifications

Accredited specializations in risk management, data analysis, computer architectures, and network protocols.

Risk & FinanceMar '26

Risk Management Specialization

Coursera

VerifiedCoursera
Data ScienceMay '25

The Complete Data Analyst Bootcamp

Udemy

VerifiedUdemy
NetworkingOct '24

Computer Communications

Coursera

VerifiedCoursera
Computer SystemsSep '24

Introduction to Hardware and Operating Systems

Coursera

VerifiedCoursera
NetworkingSep '24

The Bits and Bytes of Computer Networking

Coursera

VerifiedCoursera
AnalyticsApr '24

Excel Dashboard for Beginners

Simplilearn

VerifiedSimplilearn
Python & DataFeb '24

Learning Python for Data Analysis and Visualization Ver 1

Udemy

VerifiedUdemy
• • •
• • •

Achievements & Community Impact

Algorithmic problem-solving dedication paired with civic responsibility and volunteer leadership.

LeetCode 200+Aug '26

Solved 200+ DSA Problems in Java on LeetCode

Mastered algorithmic paradigms including Dynamic Programming, Graph Theory, Trees, Recursion, Two-Pointers, and Binary Search with optimized space-time complexities.

LeetCode Profile VerifiedDocumented in CV
Mother Teresa SansthaJun '24

Community Outreach & Social Impact Volunteering

Played a key role in delivering healthcare access, nutritional supplies, and educational programs to underprivileged communities with Mother Teresa Sanstha.

Mother Teresa Sanstha OutreachDocumented in CV
• • •
• • •

Let's Connect & Collaborate

Open for high-impact AI Engineering, Machine Learning, Data Science, and Risk Consulting roles.

Direct Communication

Whether you want to discuss Agentic AI systems, explainable machine learning models, quantitative risk assessment, or career opportunities, feel free to reach out.

Mobile / WhatsApp
+91-9709678176
LinkedIn ProfileGitHub Profile
Curriculum Vitae

Download verified credentials, ML architectures, and quantitative risk projects.

Download Resume

Send a Message

Have an inquiry, project, or open role? Send your details below.