LANGCHAIN · MCP · OPENBB
TapeDeck
A personal equity research terminal. A LangChain agent drives 127 live financial tools through the OpenBB MCP server, falling back across data providers and eight LLMs so long research runs finish.
01
About
I'm in my final semester of a Master of Engineering in Information Systems Security at Concordia University in Montréal, with a 3.97/4.30 GPA, and I'm enthusiastic about generative AI and cybersecurity.
In fall 2025 I interned as a Junior Security Operations Analyst at TMX Group, the operator of the Toronto Stock Exchange, working with enterprise-grade security tools to monitor, investigate and strengthen the security posture of critical financial market infrastructure.
I've also interned at American Express twice. In summer 2025 I interned as an Information Security Analyst on the Generative AI Security team, working hands-on with LLMOps, secure AI/ML pipeline integration and risk mitigation across production systems. In summer 2026 I returned to work on identity and access management workflows and build agentic workflows on policy engines, so AI actions stay stable and deterministic.
Outside of tech, I love reading and writing books, travelling (India, Canada, the UAE, Mexico and Peru so far) and learning new languages. I'm currently learning Spanish for better mobility. I'm writing my first novel, One Night in Trombay, and publishing articles on X and Medium.
- Based in
- Montréal, Canada
- Studying
- M.Eng, Information Systems Security
- GPA
- 3.97 / 4.30
- Writing
- One Night in Trombay
Toolkit
Focus
- GenAI security
- Security governance
- Threat intelligence & incident response
- Key risk indicators
- Emerging technology risk
Practice
- Vulnerability management
- Security investigations
- Risk reporting
- Social engineering assessments
- OSINT
Tools
- MCP
- LangChain
- LangGraph
- Splunk
- Jira
- Confluence
- Cloud security tools
- Excel (advanced)
Code
- Python
- SQL
- Bash
- JavaScript
Systems
- Windows
- macOS
- Linux
02
Work
-
American Express
Intern I
- Studied the company's existing identity and access management workflows and documented them for use across the enterprise.
- Analyzed policy engines and built agentic workflows integrated with them, so agentic AI actions stay stable and deterministic.
Information Security Analyst Intern I
- Worked in the Generative AI Security team on LLMOps, secure AI/ML pipeline integration and risk mitigation across production systems.
- Delivered cybersecurity initiatives across threat intelligence, incident response, data loss prevention, identity and access management, and cryptography services.
- Led research and development on post-quantum cryptography and GenAI security enablement.
-
TMX Group
Jr. Security Operations Analyst Intern
- Worked with enterprise-grade security tools to monitor, investigate and strengthen the security posture of critical financial market infrastructure.
- Designed and ran context-aware social engineering campaigns, including whaling and credential harvesting, to test and strengthen the human layer of defence against financial-sector threats.
- Analyzed key risk indicators to find vulnerability patterns among the C-suite and critical staff, then rolled out targeted training that reduced credential compromise rates.
03
Projects
RAG · FAISS · MITRE ATT&CK
TacTracer
An autonomous IoT threat hunter. Upload network logs and get a forensic report grounded in MITRE ATT&CK: guardrails reject anything that isn't a log, FAISS retrieves the matching techniques, and Llama 3.3 70B writes the findings. Built for IoT Security at Concordia.
GENAI · ZXCVBN · LLM FAILOVER
SentinelXC
A password architect and auditor. It turns a theme into a strong, memorable passphrase, then audits passwords with zxcvbn and an LLM red-team persona that roasts weak choices. A model ladder of Gemini, Llama 3 and Mistral keeps it running when a provider fails.
GEMINI · LLM ENSEMBLE · STREAMLIT
Nuestra Bóveda
A private AI time capsule disguised as a time-zone dashboard. Behind a hidden date-key unlock sit a memory oracle that picks a shared memory to match your mood, and a ghost writer that learns a person's tone from old chat logs.
MISTRALAI · IMAGE GENERATION
Whimsical
A children's storybook generator. MistralAI writes the story, AI image generation illustrates every page, and a character-memory system keeps the characters looking the same from page to page.
NLP · KNOWLEDGE GRAPHS
Knowledge Tree
Paste any text and get a knowledge graph. It summarizes the input, extracts factual triples chunk by chunk, and draws how everything connects.
04
Research & writing
Publications
-
PAPER · WCCIT 2023 · BERLIN
AI-Driven Threat Intelligence: Leveraging Machine Learning to Empower Cybersecurity Applications for Enhanced Threat Detection and Response
How machine learning can strengthen threat intelligence, from detection through response. Presented at the World Conference on Computing and IT in Berlin, Germany, on August 12, 2023.
-
PAPER · 16TH SINCONF · JAIPUR
Mitigating Advanced Persistent Threats (APTs) in Cybersecurity Through the Implementation of Distributed Ledger Technology
Proposes distributed ledger technology as a way to counter advanced persistent threats and make systems more resilient to intrusion. Shortlisted for presentation at the 16th SINCONF in Jaipur, India.
-
REVIEW PAPER · 2023
Analyzing Modern Cryptography Techniques and Reviewing their Timeline
Traces ciphers in chronological order, from the earliest known cryptography to modern cryptographic systems.
-
REVIEW PAPER · 2023
Process Scheduling in Operating Systems and Evolution of Windows
How operating systems schedule processes, with a focus on how Windows implements it.
Latest essays
05
One Night in Trombay
A DEBUT NOVEL · COMING SOON
A gripping tale of mystery and resilience, set against the vibrant backdrop of Bombay.
“Step into a never-ending chase.”
