Pallavi Pandey
Full Stack Engineer building the future with Generative AI, LLMs, and intelligent Data Systems.
About Me
I am a Full Stack Engineer at Dom Ventas and a Data Science graduate from IIT Madras. I specialize in building intelligent systems that merge high-performance backend engineering with state-of-the-art AI.
My primary focus is on Generative AI, where I build robust pipelines utilizing LLMs, RAG, and Agentic Workflows. I thrive on migrating complex architectures and integrating AI functionalities like TTS, STT, and translation into real-world, scalable platforms.
I am also deeply interested in Blockchain & Web3, exploring smart contract development on Ethereum & Solana, DeFi protocols, and the convergence of AI with decentralized systems to build trustless, verifiable applications.
Beyond my professional work, I am a passionate hackathon enthusiast, participating in events like the Kharagpur Data Science Hackathon 2024. I am dedicated to pushing the boundaries of what is possible with data and machine intelligence.
Education
B.S in Data Science & Applications, IIT Madras
Interests
AI Research, Blockchain & Web3, Scalable Systems, Open Source, Hackathons
Professional Experience
Full Stack Engineer
Dom Ventas, Bengaluru
- •Managing and automating core application workflows supporting operations, retail, marketing, and sales.
- •Architecting an accreditation application with a RAG-based AI chatbot for intelligent document retrieval and user assistance.
- •Maintaining a custom dashboard for unified business insights via ETL processes across multiple databases.
Software Developer Engineer
Articence Inc, CA
- •Integrated advanced Gen-AI capabilities (LLMs, RAG, TTS, STT) to build intelligent automation and real-time conversational systems across voice and chat channels.
- •Architected and migrated backend systems to FastAPI to support low-latency, high-throughput AI inference pipelines at scale.
- •Implemented real-time communication using WebSockets and Webhooks to enable live streaming interactions and event-driven workflows.
- •Contributed to scalable contact-center and automation infrastructure, improving system reliability, performance, and user experience.
Python & AI Developer
Magizh Technologies, Coimbatore
- •Designed and built an AI-powered chatbot that delivers personalized educational recommendations based on student profiles and learning goals.
- •Integrated AI to automatically generate and format quiz questions from book excerpts, content, or paragraphs, simplifying question creation and upload.
- •Developed a quiz application enabling students to add their own questions and revise them later through interactive quiz and fill-in-the-blanks formats for rapid practice.
- •Engineered scalable backend services using Flask and PostgreSQL to efficiently process and manage student data.
Junior Developer
VaayuSastra, Chennai
- •Built a learning platform where kids can learn, access, and take tests related to aerospace content.
- •Designed interactive assessment modules to track and evaluate student progress effectively.
- •Integrated educational resources with a focus on aerospace engineering concepts for high engagement.
Web Developer
IIT Madras, Chennai
- •Developed the official website for 'Paradox', the flagship technical and cultural fest of the IITM BS degree.
- •Designed and launched the project showcase website for academic and research initiatives.
- •Managed and coordinated Student Affairs, bridging the gap between student requirements and administration.
Invited Talks & Workshops
Guest Lecture: AI & Data Science
Delivered an insightful session on the evolving landscape of AI and Data Science, bridging academic theory with industry practices.
Data Analysis Training Program
Conducted a comprehensive 6-week training program on Data Analysis using SQL, Python, Pandas, NumPy, Power BI, and other modern data analysis tools, empowering students with practical industry-ready skills.
Workshop: Blockchain & Web3 Fundamentals
Conducted an intensive 10-day bootcamp on Blockchain fundamentals, covering smart contract development with Solidity, DeFi concepts, and hands-on projects deploying on Ethereum testnets using Hardhat and Web3.js.
Training: Generative AI in Practice
Delivered a hands-on 3 weeks offline intensive training for B.Tech interns at Training Institution, Bangalore Whitefield, covering LLM fundamentals, RAG pipelines, prompt engineering, agentic workflows with LangChain, and real-world deployment of AI-powered applications.
Featured Projects
Bridging the gap between robust backend engineering and data-driven intelligence.
AI Financial Modeling & Decision Intelligence
End-to-end financial modeling system with XGBoost + RandomForest revenue forecasting, what-if scenario simulation, and AI-generated business insights via a 5-provider fallback chain. Features an interactive Streamlit dashboard with Plotly visualisations.
AI Voice Assistant
Real-time AI voice agent for an auto service call center. Handles VIN lookups, customer profile creation, and call routing entirely by voice using a fully free AI stack.
AI Healthcare System
Advanced clinical documentation pipeline merging Speech-to-Text and Generative AI. Features real-time transcription and automated medical reporting using specialized LLM prompting.
AI Video Dubbing Pipeline
End-to-end intelligent video translation system. Orchestrates MarianMT for translation, OpenAI Whisper for ASR, and Wav2Lip GAN for precise lip-syncing.
IITM BS Portal Gen AI
Full-stack educational platform integrated with Google Gemini. Features AI-driven lecture summarization and automated academic question generation.
Unique Mart E-Commerce
Intelligent grocery management system with integrated sales forecasting and data visualization modules for inventory optimization.
DeFi Lending Protocol
Decentralized lending and borrowing protocol built on Ethereum. Features collateralized loans, dynamic interest rates, and on-chain liquidation logic via Solidity smart contracts.
Technical Arsenal
A comprehensive collection of tools and technologies I use to build intelligent, scalable solutions.