Experience

Where I've built things.

Roles delivering agentic AI solutions, fraud platforms, and analytics for enterprise clients.

April 2026 – Present

Forward Deployed Engineer

Peakflo

Remote

  • Own end-to-end delivery of AI Proof-of-Concepts (POCs) for enterprise finance clients, working directly with stakeholders from requirement gathering through deployment and demonstrations.
  • Design and build Agentic AI workflows using modern LLMs, implementing multi-agent orchestration, prompt engineering, tool calling, and workflow automation for finance operations.
  • Develop production-grade RAG pipelines integrating enterprise knowledge bases, emails, and business documents to improve response accuracy while minimizing hallucinations.
  • Engineer optimized prompts for finance AI agents through iterative evaluation, prompt versioning, self-reflection techniques, and A/B testing to improve quality and reliability.
  • Integrate AI workflows with backend systems using Python, MongoDB, APIs, OCR pipelines, and business automation frameworks to streamline customer workflows.
  • Collaborate cross-functionally with Product, Engineering, Customer Success, and enterprise clients, taking complete ownership of solution architecture, implementation, testing, documentation, and production rollout.

April 2025 – April 2026

Data Engineer

UNITE Hub Technologies

Chennai, Tamil Nadu, India

  • Architected and developed an AI-powered Fraud Risk Management (FRM) platform for banking institutions, successfully onboarding City Union Bank (CUB) as the first enterprise client.
  • Built an end-to-end full-stack application using FastAPI, PostgreSQL, and Vue.js for real-time fraud detection and transaction monitoring.
  • Integrated the I4C (Indian Cyber Crime Coordination Centre) API to process fraud complaints with automated RRN validation, transaction matching, and case generation.
  • Developed an AI-powered analytics interface using Ollama LLM, enabling natural language querying across banking databases for faster fraud investigations.
  • Automated email processing using n8n, extracting structured information from risk officer emails and creating fraud cases automatically, significantly reducing manual effort.
  • Deployed scalable services on Google Cloud Platform (GCP) with asynchronous processing, Keycloak authentication, and production-ready architecture supporting high transaction volumes.

June 2024 – Sept 2024

Data Analyst Intern

Dhaval Ads

Ahmedabad, Gujarat, India

  • Supported business operations through data collection, analysis, and reporting.
  • Built analytical reports and generated insights that enabled data-driven decision making.
  • Assisted in improving operational efficiency through business analytics and reporting.