Hello, I'm
Star Scholar at SAP Labs Bangalore & M.Tech at BITS Pilani. Building enterprise AI systems, cognitive research, and full-stack products.
0+
Projects Built
0+
Internships & Roles
0
Research Papers
0.0/10
GPA at KJC
Currently at
SAP Labs India
Building AI that matters
I'm V Meenakshi Iyer — a Star Scholar at SAP Labs Bangalore and M.Tech student at BITS Pilani, specializing in Software Engineering with SAP.
My work spans enterprise AI systems on SAP BTP (EAHOP), cognitive AI research (CogDriftFormer, SYNC), and full-stack products. Interned at SBARI California, IIT Bombay, and Nestlé India. Passionate about Human-Computer Interaction and Responsible AI.
Achievements
UiPath Student Developer Champion
Recognized as a leading student developer champion by UiPath for contributions in automation and robotic process automation (RPA).
Selected for LinkedIn Live
Selected for LinkedIn Live under the EmpowerHER initiative.
Cleared CA Foundation
Cleared CA Foundation in the 1st Attempt.
Galactic Impact Award Winner
Won the Galactic Impact Award at NASA Space Apps 2024 for 'EarthWise,' an SDG-focused web app driving sustainability awareness.
Contributor to GirlScript Summer of Code (GSSoC) 2024
Actively contributed to open-source projects during the GirlScript Summer of Code (GSSoC) 2024.
University Topper in Bharatnatyam
Achieved the position of university topper in Bharatnatyam from Kairagarh University.
Skills
Experience
SAP Labs
Bangalore, India
Star Scholar
Internal Teams
Responsibilities
- Strengthened API quality by expanding Mocha test coverage and improving validation logic
- Diagnosed and fixed functional issues, including a Joule scenario bug, while resolving accessibility compliance tickets
- Contributed to an AI-driven RAG, predictive modeling, knowledge graphs, and BM25-based retrieval POC for Maintenance Plan workflows
- Implemented backend versioning and helped deliver the Change Log feature end-to-end across UI, validations, and backend
Technologies
South Bay Area Research Institute
California, USA
Software Engineer Intern
- Collaborated with a global team on the SiNET project — full-stack development across UI/UX, frontend, and backend
- Built secure data model explorer for IoT and compliance; improved data processing by 25% and reduced latency by 15%
IIT Bombay
Remote
IT Project Management Intern
- Collaborated with Prof. Dr. Ganesh Ramakrishnan's CS dept on the Open Character Recognition (OCR) Project
- Handled project management, documentation, presentations, development, and testing to enhance OCR accuracy
Nestlé India
Bangalore, Remote
Research Intern
- Researched and identified AI-driven automation tools for Accounts Receivable within SAP technology
- Analyzed capabilities to enhance efficiency by 20% and delivered findings within the designated timeline
Projects
EAHOP
Production-deployed Enterprise AI Agent Harness & Optimization Platform on SAP BTP. Autonomously optimizes AI agents by analyzing uploaded codebases through 6 novel engines: UAR, Agent Genome, Digital Twin, OKG, OME, and AOS. Evaluates 22 performance metrics and delivers the winning optimized agent as a GitHub pull request. Achieved +17.6% average accuracy improvement and −11.6% API cost reduction across 5 production agents. Stack: Python/FastAPI, React 18, SAP Fiori, Claude 3.7 Sonnet via SAP AI Core. 86 automated tests passing.
CogDriftFormer
Novel multimodal transformer architecture for real-time cognitive safety monitoring. Detects cognitive overload, emotional drift, and manipulation exposure from behavioral event sequences combined with natural language signals. Introduces DriftAware Positional Encoding and Cross-Modal Dual Attention mechanisms. 21.77M trainable parameters. Key results: OL-QWK 0.882 (ordinal regression), MA-AUROC 1.000 (anomaly detection), MA-F1 0.750. IEEE conference paper under review.
SYNC
Emotionally-aware 3D avatar AI companion for first-generation college students, designed to proactively prevent academic disengagement. Tracks emotional trajectory in real time and intervenes before burnout occurs. Implements 10 novel research mechanisms: AEM, TEFF, MBDS, PantoMatrix audio-driven motion engine, and more. Stack: React, Three.js (R3F), ReadyPlayerMe avatar, Ollama Qwen2.5:7b, XTTS v2, Whisper ASR, custom motion pipeline. Research prototype under IP review.
Skylark Men's Wear
Full-stack ERP platform for a men's tailoring boutique, digitizing the complete garment lifecycle from measurement to delivery. Features 6-stage order tracking (PENDING → MEASURING → CUTTING → STITCHING → READY → DELIVERED), role-based portals for Owner/Worker/Customer, dynamic garment measurements, automated WhatsApp notifications via Baileys, loyalty points engine, invoice PDF generation, and PWA offline support. Stack: Next.js 16, Prisma 7, Neon PostgreSQL, Tailwind CSS, shadcn/ui, custom JWT auth.
i-ScOper AI Enhancement
AI-powered enhancement layer for SAP's Intelligent Test Scope Optimizer (i-ScOper). Applies an ML ensemble of Random Forest, Gradient Boosting, XGBoost, and LightGBM with Bayesian scoring and rolling 4-fold cross-validation to intelligently prioritize regression test cases. Includes a full EDA pipeline for data profiling, correlation analysis, and outlier detection. Reduces unnecessary regression cycles and accelerates release velocity for enterprise SAP application development.
AIOS
Natural-language AI Operating System for macOS that interprets voice or text commands and executes OS-level actions: launching applications, controlling browsers, organizing files, and running terminal commands. Intent engine built with spaCy covering 15+ action categories. Includes a RAG-based self-learning module for continuous improvement. LLM fallback powered by Ollama + Phi-3 Mini for complex queries. Packaged as a cross-platform Electron app with a React frontend and Flask backend.
- 07View Case Study
EarthWise
Interactive web platform teaching all 17 UN Sustainable Development Goals through AI-driven immersive lessons. Features a lifelike AI Tutor with real-time lip-sync and Text-to-Speech, multilingual support for global accessibility, and gamified learning paths. Winner of the NASA Space Apps 2024 Galactic Impact Award. Stack: Next.js, GPT-3 API, ElevenLabs TTS, Rhubarb Lip Sync, Google Translate API, Python/Express backend.
AI Emotional Intelligence Assistant
Real-time emotionally intelligent virtual assistant delivering human-like voice interaction with synchronized facial expressions and avatar animations. Implements a full multilingual speech-to-speech pipeline: ASR for speech understanding, LLM-based emotional reasoning for empathetic responses, high-quality female TTS with synchronized lip-sync output. Scalable architecture enables production-grade deployment for emotionally aware human-AI interaction scenarios.
Lecture-STT
Full-stack academic productivity tool enabling students and researchers to transcribe lecture audio, extract text from PDF slides and images, and query content using conversational AI. Achieves high transcription accuracy with Whisper ASR and supports complex PDF layouts via pdfjs-dist and Tesseract.js OCR. AI Q&A powered by Groq API (LLaMA 3.3 70B). Stack: Next.js, Node.js, Express.js, MongoDB Atlas, Multer for file uploads.
Assist — AI Voice Assistant
Python-based multilingual AI voice assistant capable of executing tasks and answering queries through natural voice commands. Integrates SpeechRecognition for input, gTTS/pyttsx3 for speech synthesis, and the OpenAI API for intelligent responses. Features a cross-platform Kivy UI for seamless use across Windows, macOS, and Linux. Supports multiple languages, enabling inclusive voice-driven human-computer interaction.
- 11View Case Study
SiNet Application
Secure IoT data governance platform for RFID-enabled circular economy operations, developed for the South Bay Area Research Institute (SBARI), Gilroy, California. Enables real-time tracking of reusable cups across sustainable transport and recycling networks. Improved data processing throughput by 25%, boosted user adoption by 30%, and reduced API latency by 15% compared to the prior system. Stack: Linux, Flask, React, REST APIs.
- 12View Case Study
ConvoBuddy
Research study deploying three conversational AI models — DialoGPT, GPT-2, and LLaMA — as Telegram chatbots to evaluate response quality, coherence, and user perception of AI-generated dialogue. Conducted structured comparative analysis measuring user interaction patterns, engagement metrics, and model performance. Provides empirical insights for future conversational AI system design and LLM selection criteria.
- 13View Case Study
AutoQuestionGen
Intelligent question paper generation system powered by UiPath RPA automation. Accepts curriculum inputs, desired difficulty distribution, and topic coverage parameters, then autonomously generates customized academic assessment papers. Reduced manual preparation time by 40% and eliminated human formatting errors. Integrates AI-assisted content selection with RPA workflow orchestration to streamline educational assessment at scale.
- 14View Case Study
Timetable Automation
Python web application automating academic timetable scheduling for educational institutions. Handles constraint satisfaction across classes, instructors, and room availability while minimizing conflicts and manual coordination overhead. Exposes a RESTful backend API for programmatic timetable generation and a responsive frontend for visualization, editing, and export. Eliminates scheduling errors and saves administrative hours each academic cycle.
- 15View Case Study
Attendance Application
Native Android application automating student attendance marking with an intuitive mobile interface optimized for speed and accuracy. Replaces manual roll-call processes, improving attendance tracking accuracy by 40% and cutting administrative time significantly. Features real-time data sync, report generation, and a clean UX designed for both faculty and institution administrators. Built with Android SDK, Java, and Firebase backend.
Presented Papers
CRYPTOGRAPHY AS A PART OF E-COMMERCE RECORD MANAGEMENT SYSTEM
Examines the role of cryptography in securing e-commerce transactions. Discusses double-entry bookkeeping in e-commerce, explores security threats like phishing and SQL injections, and explains cryptographic techniques including encryption and hashing. Highlights SHA-256 as the most secure hashing function for protecting e-commerce transactions and preventing data breaches.
ICCI202459: COMPREHENSIVE ANALYSIS OF ARTIFICIAL INTELLIGENCE IN CONTEMPORARY SOCIETY
Analyzes the current state of AI, focusing on transformative applications and societal implications. Presents a comparative study of methodologies to enhance AI utility, addressing ethical concerns like bias and job displacement while emphasizing responsible development. Highlights AI's potential to revolutionize education and customer service.
ENHANCING STUDENT ENGAGEMENT WITH SDGs THROUGH GAMIFICATION AND AI IN THE LEARNING ENVIRONMENT
Co-authored study exploring how AI-powered tools and gamification increase student engagement and motivation toward Sustainable Development Goals. Demonstrates that gamified challenges, simulations, and AI-driven personalized learning enhance understanding and foster responsible global citizenship. Provides actionable frameworks for integrating AI and gamification into classroom teaching.
Contact Me
Let's Connect
I'm always open to new opportunities, research collaborations, and interesting conversations. Whether you have a question, a project idea, or just want to say hi — my inbox is always open!
