Anshuman Bakshi

About Anshuman Bakshi

I am a Software Engineer and Researcher specializing in the high-impact intersection of Gen-AI + Cybersecurity + FinTech. My expertise lies in building secure, privacy-preserving intelligent systems that bridge the gap between technical architecture and complex business logic.Experience & Impact:National University of Singapore (NUS): Recently completed a Research Fellowship focused on Applied ML and GenAI. I conducted feasibility studies that identified automation workflows to reduce operational overhead by 30% and optimized AI deployment speeds by 15% using Agile methodologies.Panasonic India: As a Software Engineering Intern for the Miraie Platform, I performed deep Gap Analysis on testing workflows. I implemented REST API automation using Go, which resulted in a 45% increase in operational efficiency and streamlined the bridge between technical development and business requirements.Flagship Major Project (SRMIST): SentinXFL For my major project at SRM, I architected SentinXFL, a federated fraud detection framework. By implementing Differential Privacy (ε=4.5), I balanced technical security with business data-sharing needs, ensuring the system meets strict PCI-DSS and GDPR compliance standards while maintaining a 93% accuracy rate in detecting emerging fraud trends.Core Competencies:AI & Cybersecurity: Generative AI, Federated Learning, Cryptography, Privacy-by-Design, Zero-Trust Architecture.Business Analysis: Requirements Gathering, Gap Analysis, ROI Analysis, and Stakeholder Management.Technical Stack: Go (Backend), Python, Java, SQL, TensorFlow, Apache Kafka, and REST API Automation.I am currently seeking Research Fellowships and Software Engineering roles where I can leverage my experience in secure AI and distributed systems to drive measurable business value.Let’s connect to discuss the future of secure FinTech and AI.

Experience

1. Generative Ai Engineer Intern

   GlobalLogic

   Jan 2026 — Present · Noida, IN

   Building MCP-based workflow services that enable Codebuddy to trigger multi-step agentic workflows (plan → execute → validate) with stable tool contracts.Designing security-by-default execution: default-deny policies, allowlists, sandboxed outputs, and structured audit events for every tool-like action.Implementing an evidence pipeline that writes deterministic run artifacts (manifest/events/report/policy snapshot/artifact index) to support debugging, governance, and review.Partnering with the team on roadmap exploration for Physical AI, including early prototyping considerations for drone-focused agentic automation (restricted tool access, deterministic validation).

Education

• Shalom Hills International School

  High School/Secondary Diplomas and Certificates

• SRM IST Chennai

  Bachelor of Technology - BTech, Computer Science