Available for opportunities

Swastik
Ghosh

Manager at HSBC Financial Crime Analytics.
Engineer by training, quantitative strategist by practice —
building at the intersection of Mathematical Finance, Multi-Agent AI, and Algorithmic Intelligence.

View Resume
⚙️ Technical
97.5%
Leakage Detection Accuracy
SVM, XGBoost & Linear Regression on multi-sensor industrial data at Hitachi
Hitachi · 2023
🤖 Technical
60%
Automation Gain via LLMs
End-to-end LLM pipeline: code parsing, embedding graphs, RAG SQL & vector DB at EY
Ernst & Young · 2024
🛡️ Technical
40%
Better Threat Detection
LLM-powered vulnerability scanner using transformer NLP at EY
Ernst & Young · 2025
🔬 Technical
45%
AI Accuracy Boost
Responsible AI models for toxicity/bias detection — significant NLP safety benchmark improvements
Ernst & Young · 2025
💉 Technical
50%
Faster Security Scans
SQL injection testing automation slashed scan time & reduced manual effort 60%
Ernst & Young · 2025
🏆 Leadership
Manager
Global Financial Crime Lead
Overseeing case prioritisation & allocation across Singapore, UK, and Turkey markets
HSBC · 2026
📚 Publication
IJFMR
Published Research · Fraud Detection
Multi-Model Financial Fraud Detection with Explainable AI — ensemble ML for compliance teams
IJFMR · 2025
📖 Publication
IJSREM
Published Research · Blockchain ML
ML-Based Spectrum Sensing & Management using Blockchain for distributed networks
IJSREM · 2023
🎓 Mentorship
Volunteer
Teaching & Peer Mentorship
Led volunteer teaching in data science fundamentals, supporting junior peers on real projects
Community · Ongoing
📜 Certification
LBS
Financial Analysis · London Business School
Financial modelling, valuation frameworks, and quantitative analysis methodologies via Coursera
London Business School · Coursera
2+ Years Experience
2 Companies
2 Publications
60% Avg. Efficiency Gain
Career

Work Experience

Click on a company to explore the full details →

🏦
HSBC
Manager
Feb 2026 — Present
↗
📊
Ernst & Young
Consultant → Assoc. Consultant
Aug 2024 — Feb 2026
↗
⚡
Hitachi
Research Intern · Data Science
Aug 2023 — Nov 2023
↗
Stack & Expertise

Technical & Quant Skills

Programming & Tools
Python
R
C++
SQL
MATLAB
Git
PySpark
Flask
VS Code
Postman
Quantitative Finance
Probability
Statistics
Time Series Analysis
Stochastic Processes
Optimization
Numerical Methods
Monte Carlo Simulation
Financial Skills
Financial Modelling & Valuation
Financial Analysis
Derivatives Pricing
Black-Scholes
Options
Market Microstructure
Algorithmic Trading
Market Impact
Risk Management
Portfolio Analysis
ML & AI Expertise
Machine Learning
Deep Learning
NLP
LLMs
Generative AI
Databases & Visualization
MongoDB
MySQL
Tableau
Power BI
Platforms
AWS
Figma
Built

Projects

01
Finance AI Agent
Python · HTML/CSS · Flask · Finhub API
AI assistant delivering real-time stock updates, financial insights, and market history with personalised recommendations. Implements trend analysis and stock suggestion logic.
View Project
02
Sentiment Analysis Chrome Extension
Python · JavaScript · HTML/CSS · Flask
Identifies sentiment keywords, computes text polarity, and visualises results through charts and word clouds. Packaged as a Chrome extension for real-time text analysis.
View Project
03
Monte Carlo Wealth Simulator
Finance · Simulation · Python
Simulates wealth trajectories using Monte Carlo methods to model various market conditions and investment strategies over time.
View Project
04
Market Impact & Execution Cost Simulator
Finance · Trading · Python
Models the market impact and execution costs for large algorithmic trades, simulating slippage and liquidity consumption.
View Project
05
Order Book Liquidity Engine
Finance · Market Microstructure · Python
Simulation engine for order book liquidity, matching bids and asks dynamically to evaluate market microstructure scenarios.
View Project
06
Algorithmic Trading Battle Simulator
Finance · Algorithmic Trading · Python
Platform to backtest and pit different algorithmic trading strategies against each other in historical market environments.
View Project
07
Black-Scholes Visual Explainer
Finance · Derivatives · Data Viz
Interactive visual explainer for the Black-Scholes options pricing model, breaking down Greeks and volatility impact dynamically.
View Project
08
Volatility Surface Builder
Finance · Quantitative Analysis · Python
Constructs and visualises implied volatility surfaces for options chains across various strikes and maturities.
View Project
09
Systemic Risk Contagion Engine
Finance · Risk Management · Network Theory
Models systemic risk and financial contagion across interconnected banking networks to assess cascading defaults.
View Project
Research

Publications

2025
Multi-Model Financial Fraud Detection with Explainable AI
IJFMR — International Journal For Multidisciplinary Research
Ensemble approach to financial fraud detection incorporating Explainable AI (XAI) to make model decisions interpretable for compliance and audit teams. Covers feature importance, SHAP analysis, and model benchmarking.
2023
ML-Based Spectrum Sensing and Management Using Blockchain
IJSREM — International Journal of Scientific Research in Engineering and Management
Machine learning applied to dynamic spectrum sensing in cognitive radio networks, combined with blockchain for secure, tamper-proof spectrum allocation and management across distributed nodes.

Let's Connect

Open to roles in Quantitative Finance, Financial AI, Quant Development, and Risk Analytics. Always happy to talk.

✉ swastikghosh1106@gmail.com 🔗 linkedin.com/in/swastik-ghosh ⌥ github.com/sg1106 📞 +91 91312 30644
Skill Alignment

Quant & Financial AI

A practitioner operating across Quantitative Finance and Financial AI Engineering — combining mathematical modeling with production-grade AI systems.

01
📈
Quantitative Finance
From quantitative financial crime detection at HSBC to probabilistic risk modeling at EY, I apply mathematical finance, Monte Carlo simulations, and derivatives pricing frameworks to high-consequence financial environments.
Quantitative Risk Monte Carlo Derivatives Volatility Surface AML Market Microstructure
Explore Alignment →
02
⚙️
Financial AI & Engineering
Deep engineering experience in Multi-Agentic GenAI, Text-to-SQL for financial databases, RAG compliance systems, PyTorch, and automated web-scraping intelligence for global markets.
Multi-Agent GenAI Text-to-SQL Financial RAG PyTorch / TF Web Intelligence Algorithmic Engines
Explore Alignment →
⬡ Skill Alignment · Finance

Finance

From detecting financial crime at a global bank to building probabilistic risk quantification models — my work lives at the intersection of data, money, and consequence.

Current Role
Financial Crime Detection · HSBC
Leading automated web-scraping systems to identify financial crime for individuals across Singapore, UK, and Turkey. Designing trigger generation for suspicious transaction detection.
3 Markets
AML Domain
EY Consulting
CyberRisk Quantification System
Built a quantitative risk system using probabilistic models, Monte Carlo simulations, and SQL-based threat intelligence to generate risk-adjusted financial mitigations.
45% Efficiency Gain
Monte Carlo Method
Project
Finance AI Agent
Personal AI assistant for real-time stock updates, financial insights, market history, and personalised investment recommendations using the Finhub API and trend analysis logic.
Publication
Multi-Model Financial Fraud Detection with XAI
Published research on ensemble approaches to financial fraud detection, incorporating Explainable AI to make model decisions interpretable for compliance and audit teams.
Certification
Fundamentals of Financial Analysis
London Business School via Coursera — covering financial modelling, valuation frameworks, and quantitative analysis methodologies.
Certification
SEBI Investor Awareness Test
Passed the SEBI/NISM Investor Awareness certification — demonstrating foundational knowledge of Indian securities markets, regulations, and investment principles.

Finance Journey

2023
Predictive Maintenance Modelling · Hitachi
Applied ML to industrial sensor data — an early application of quantitative modelling with real-world consequences: accuracy, downtime reduction, and cost impact.
2024–2025
CyberRisk Quantification · EY
Built probabilistic financial risk models with Monte Carlo simulations. First direct application of financial modelling in a professional consulting context.
2025
Fraud Detection Publication · IJFMR
Published research on financial fraud detection using multi-model ML and Explainable AI. Connected research to real financial compliance needs.
2026 →
Manager, Financial Crime · HSBC
Now operating at the managerial level within a Tier-1 global bank, directly combating financial crime across multiple international markets.
⬡ Skill Alignment · Technology

Technology

I build production-grade ML and AI systems with measurable outcomes. My engineering spans from transformer-based NLP to end-to-end LLM pipelines — always in service of real business problems.

EY · LLM Engineering
LLM Feature Suite in Python
Developed full LLM pipeline: code parsing, embedding-based graph generation, automated PDF/Excel output, and RAG SQL/vector DB integration using PyTorch and TensorFlow.
60% Automation Gain
EY · Security AI
LLM-Powered Vulnerability Scanner
Built transformer-based NLP system with conversational interface for real-time vulnerability detection. Improved threat detection by 40% over baseline methods.
40% Better Detection
EY · Responsible AI
Bias & Toxicity Classification Models
Built and evaluated Responsible AI models to detect and classify toxicity, bias, and unsafe language — 45% accuracy boost and significantly improved NLP safety benchmarks.
45% Accuracy Gain
EY · Security Automation
SQL Injection Testing Automation
Combined web crawling, ML, and GenAI payload generation to automate SQL injection testing for DVWA. Slashed manual effort by 60% and scan time by 50%.
60% Effort Reduction
50% Faster Scans
Hitachi · ML Research
Leakage Detection with Sensor Data
SVM, XGBoost, and Linear Regression models integrating pressure/flow/temperature sensors with system topology. Achieved 97.5% detection accuracy.
97.5% Accuracy
Publication
Spectrum Sensing via ML + Blockchain
Published in IJSREM — ML-based spectrum sensing and management using blockchain, demonstrating early research capability across wireless systems and distributed ledger technology.

Technology Journey

2023
First ML Models · Hitachi Research Internship
Built SVM, XGBoost, and regression models on industrial sensor data — first exposure to production ML with real accuracy constraints (97.5%). Learned to work with noisy, multi-modal data.
2023
Research Publication · IJSREM
Published ML-based spectrum sensing using blockchain. Demonstrated ability to apply ML to novel, cross-domain problems early in the career.
2024
LLM Engineering at Scale · EY
Built end-to-end LLM pipelines with PyTorch/TensorFlow including code parsers, embedding-based graph generation, RAG integrations, and automated document output — 60% automation gain.
2025
Security AI & Responsible AI · EY
Shipped transformer-based vulnerability scanners (40% better detection) and Responsible AI classifiers for bias/toxicity (45% accuracy gain). Also automated SQL injection testing — cutting scan time 50%.
2026 →
AI-Powered Financial Crime Detection · HSBC
Applying ML and web-scraping at global banking scale — building systems that identify financial crime across international markets using name-based intelligence and pattern recognition.
Career Journey

Work Experience

3+ years across global banking, top-tier consulting, and applied research — building real systems with measurable impact.

🏦
HSBC · Financial Crime & Future Analytics
Manager
Feb 2026 — Present · Bangalore, India
Current
🤖
Multi-Agentic GenAI EWRA Analytics
Built a multi-agentic GenAI system for EWRA financial-crime analytics, directly connecting to a multi-table database and using GenAI models as tools for schema-aware database understanding, user-query intent classification, and Text-to-SQL generation, followed by automated query execution, data retrieval, visualization, and client-ready PDF report generation.
📑
Multi-Agentic RAG for EWRA Reports
Developed a multi-agentic RAG system to retrieve relevant data from EWRA reports and perform mathematical operations and contextual analysis using GenAI models as tools.
🕸️
Automated Web-Scraping Intelligence
Designed and developed an automated web-scraping system that aggregates publicly available internet data to identify financial crime and financial activity related to individuals using name-based searches.
📋
Case Prioritisation & Workflows
Case prioritisation and allocation workflows across Singapore, UK, and Turkey markets — coordinating multi-team dependencies and suspicious transaction triggers in real time.
3 Markets
GenAI Multi-Agent
EWRA Analytics
📊
Ernst & Young · Data Science & ML
Consultant → Associate Consultant
Aug 2024 — Feb 2026 · Bangalore, India
🤖
LLM Feature Pipeline
Code parsing, embedding-based graph generation, PDF/Excel output and RAG SQL/vector DB integration — 60% automation gain.
60% Automation ↑
🛡️
LLM Vulnerability Scanner
Transformer-based NLP system with conversational interface for real-time threat detection. 40% better than baseline methods.
40% Better Detection
⚖️
Responsible AI Models
Toxicity, bias, and unsafe language classification models with 45% accuracy improvement over baseline safety benchmarks.
45% Accuracy ↑
💉
SQL Injection Automation
Combined web crawling, ML & GenAI payload generation to automate SQL injection testing. Slashed manual effort 60%, scan time 50%.
60% Effort ↓ · 50% Faster
📉
CyberRisk Quantification
Probabilistic risk models using Monte Carlo simulations and SQL threat intelligence for risk-adjusted financial mitigations. 45% efficiency increase.
45% Efficiency ↑
60% Automation
40% Better Detection
1.5 yrs Duration
⚡
Hitachi · Research Intern
Data Science & Machine Learning
Aug 2023 — Nov 2023 · Bangalore, India
🔍
Leakage Detection ML Models
SVM, XGBoost, and Linear Regression models integrating pressure/flow/temperature sensors with system topology. Achieved 97.5% detection accuracy.
97.5% Accuracy
⚡
Performance Improvements
Improved response time by 35% and lowered false positives by 60% through model optimisation and feature engineering.
35% Faster · 60% Fewer FP
97.5% Accuracy
35% Faster
ML Research

Resume LLM

Online (Gemini + RAG)

Hello! I am Swastik's Resume AI Assistant. I can answer questions about his experience, skills, publications, and projects based on his resume. What would you like to know?