Overview
Summary
Work History
Education
Skills
Websites
PUBLICATIONS & COMMUNICATION
Projects
Timeline
bgImages
ARZOO RAI
AI Tech Intern, Marketing & Commerce

ARZOO RAI

Deloitte, Technology & Transformation
Delhi

Overview

Computer Science undergraduate (2027) who takes ambiguous technical questions and turns them into structured, measurable answers leadership can act on. Currently in Deloitte's Technology & Transformation practice, Marketing & Commerce, shipping products to global clients across regions. Founder of an early-stage venture built around a gap I identified in India's startup ecosystem, author of two peer-reviewed publications, and builder of an open-source platform grown from a personal repository to 29 contributors. Curious about AI and the future of work, and comfortable working independently when the guidance runs out.
01+
Year of experience

Work History

06.2026 - Current5 Months
AI Tech Intern, Marketing & Commerce
Deloitte, Technology & Transformation
  • Work across functions and regions inside the marketing and commerce technology practice, delivering products for global enterprise clients.
  • Took an open-ended question, whether an AI voice product was good enough to release, and built the structured evaluation framework that answered it: a 2,880-run test matrix across 47 voices and 12 intents per language in English, Chinese and Hindi, where previously there was no agreed way to measure it at all.
  • Translated the findings into a recommendation for leadership, quantifying a false-trigger rate of 377 per hour against a target of under 0.5, corroborating it independently against real-world audio, and setting out exactly which gaps needed closing before launch.
  • Audited the framework itself before trusting its output and found 12 defects, 3 of which had produced numbers that would otherwise have been reported as real findings.
  • Coordinated delivery of a second product handed over to the China team, working across geographies and time zones.
06.2025 - 08.20252 Months
Intern
Univest, Technology Intern
  • Built and shipped UniBot, an AI assistant answering customer questions inside a live fintech product, owning it end to end against a live user base rather than a sandbox.

Education

Bachelor of Technology | Computer Science (AI Specialization)

Bennett University | 05.2027
Bennett University
05.2027
  • AWS Certified Machine Learning Engineer, Associate.
  • GPA: SGPA 8.68 / 10

Skills

Analysis & Problem Solving: breaking ambiguous problems into structured approaches
business and market research
competitive analysis
framework building
controlled comparison and ablation studies
hypothesis testing
growth and whitespace identification
operational improvement

PUBLICATIONS & COMMUNICATION

  • Two peer-reviewed publications: SCAATA, accepted at IEEE FMLS 2026 (Japan), and Volatility-Adaptive Candlestick Pattern Recognition at ICCEt 2026. Both required a precise written argument to an audience that could check every claim.
  • Published explainers on AI for non-specialist readers, on agentic AI for autonomous trading and on graph neural networks for weather forecasting. [Articles]
  • Workshop delivered to 150+ participants on workflow automation as Open Source Lead at Google Developer Group. Mentored 15+ contributors and reviewed 30+ pull requests at GirlScript Summer of Code, ranked top 50 of 300+ mentors nationally.

Smart India Hackathon: ranked 26 of 500+ teams. 442 problems solved across competitive programming platforms. [Profile]

Projects

  • SCAATA: Self-Critiquing AI Trading Agent (accepted, IEEE FMLS 2026) Write-up Paper, First research paper I wrote, taken from an unstructured idea to an accepted IEEE submission largely independently, after the guidance available to me ran out., Used LLMs to mine, normalize and categorize 79 heterogeneous trading strategies into one structured corpus, then built an agent that critiques its own decisions., Ran a 5-config ablation to isolate what each component actually contributed rather than reporting a single headline figure, and scoped the paper's claims to what the evidence supported.
  • Founder of Founders (Founder, early-stage venture), Identified a gap in India's startup ecosystem and built a venture around it: a verified network for founders and the funders who back them, resting on a two-sided accountability model where founders rate funders on responsiveness, transparency and value-add, inverting a dynamic in which only one side is ever evaluated., Scoped a Launchpad flow turning Indian incorporation and compliance (DPIIT, INC-20A, DIR-3 KYC) from scattered bureaucracy into a tracked, deadline-aware checklist, after mapping where first-time founders actually get stuck., Produced the full v0.1 product definition and design system solo: 7 end-to-end screens, a semantic colour system, component states, dark mode and accessibility rules.
  • Eco-Play: Learning Platform (grown from a personal repo to a live community) Code, Grew a personal project into a platform selected for GirlScript Summer of Code 2026, now at 29 contributors, 54 forks and 7 mentors, by rebuilding the backlog so new contributors could onboard without needing me.

Timeline

AI Tech Intern, Marketing & Commerce
Deloitte, Technology & Transformation
06.2026 - CurrentRead more
Intern
Univest, Technology Intern
06.2025 - 08.2025Read more
Bennett University - Bachelor of Technology, Computer Science (AI Specialization)
Read more
ARZOO RAI