Hi, I'm
Mallikarjuna
Reddy Gayam
AI Engineer with 3+ years designing production-grade LLM systems, RAG pipelines, and MLOps infrastructure. Currently at Zoom, optimizing inference latency and building domain-specific fine-tuned models at scale.
// AI Engineer @ Zoom
const engineer = {
name: 'Mallikarjuna Reddy Gayam',
role: 'AI Engineer',
company: 'Zoom',
focus: [
'LLM Inference',
'RAG Pipelines',
'MLOps',
'Fine-Tuning',
],
}
export default engineer;
Professional Journey
Building production-grade AI systems and full-stack applications across enterprise environments
AI Engineer
- Deployed production LLM inference pipelines on AWS EKS using vLLM, improving throughput through dynamic batching and prompt caching strategies.
- Architected end-to-end RAG systems using LangGraph and Pinecone — owning the full pipeline from embedding model selection and chunking strategy to reranking and retrieval quality evaluation with RAGAS.
- Built MLOps automation pipelines with MLflow and GitHub Actions, enabling prompt regression testing on every code merge and proactive monitoring of embedding drift.
- Fine-tuned domain-specific LLMs using LoRA and PEFT on AWS SageMaker, validated through LLM-as-judge evaluation pipelines.
- Contributed to cross-functional architecture discussions on model deployment strategies, inference optimization tradeoffs, and production readiness criteria for generative AI features.
Software Engineer
- Developed high-performance SPAs using React.js, Next.js, and Node.js, reducing user bounce rate by 30% through optimized client-side routing and lazy loading.
- Designed and deployed TensorFlow-based ML models to automate fraud detection and customer support workflows in enterprise finance and e-commerce environments.
- Built secure, token-based RESTful APIs with Express.js, implementing input validation and role-based access control across large-scale user bases.
- Containerized microservices using Docker and deployed on AWS Lambda and Kubernetes with auto-scaling configurations and CI/CD pipelines.
Software Engineer Intern
- Built an internal operations dashboard using React and FastAPI, streamlining record tracking workflows and reducing administrative overhead.
- Integrated NLP-powered search using spaCy, Scikit-learn, and fuzzy matching algorithms, improving retrieval precision over semi-structured enterprise datasets.
- Developed secure backend APIs with JWT authentication and role-based access control, ensuring controlled data access across multiple user types.
- Deployed backend services on AWS Lambda with DynamoDB, reducing infrastructure costs while maintaining high availability.
Technical Toolkit
Across AI/ML, full-stack engineering, and cloud infrastructure
Languages & Frameworks
AI / ML & NLP
MLOps & Infrastructure
Data & Storage
Cloud & DevOps
Featured Work
Production-grade applications at the intersection of AI and full-stack engineering
Acco Finder – AI-Powered Housing Platform
Housing discovery platform that cuts student search time by 50% with ML-powered recommendations and real-time map search.
- Built with React, Next.js, and MongoDB; cosine-similarity recommendation engine using pandas.
- Real-time chat, map-based search, and Vercel serverless deployment with performance-optimized caching.
AI-Powered Resume & Cover Letter Generator
Document generation platform with 93% reported ATS success rate, powered by GPT-4 and containerized for scale.
- OpenAI GPT-4 for content generation with prompt engineering and semantic alignment to job descriptions.
- Selenium Chrome headless for pixel-perfect PDF rendering; Docker + Firestore for persistent session management.
Entry-Level Jobs Dashboard
Full-stack job discovery platform aggregating listings from LinkedIn, Indeed, and Google Jobs with AI-driven resume matching.
- TF-IDF + XGBoost for entry-level classification; personalized job scores and gap analysis.
- Interactive analytics dashboards (Chart.js / Recharts) with real-time Firestore data and Firebase Auth.
AI Article Summarizer
Real-time AI summarization tool built on GPT-4 with smart caching achieving 70% faster load times.
- Next.js frontend with TypeScript; Node.js backend APIs for efficient OpenAI model integration.
- Smart caching strategy and optimized database interactions slashed loading speed by 70%.
Portfolio Website
This portfolio — a responsive, SEO-optimized Next.js site with dynamic animations and dark-mode-first design.
- Built with Next.js, Tailwind CSS, and Framer Motion for smooth scroll-triggered animations.
- Fully responsive with performance optimizations, accessibility, and real-time project previews.
Emotion-Based Music Player
CNN-powered music recommendation system that analyzes live facial expressions to suggest personalized tracks.
- Improved user engagement by 40% through real-time mood detection and personalized playlist curation.
- Published research presented at ICSCDS 2022.
Academic Background
Master of Science in Information Systems
Saint Louis University
GPA: 3.9 / 4.0Bachelor of Science in Computer Science
Lakireddy Bali Reddy College of Engineering