Full-Stack SDE & AI Engineer | Built RAG systems in production (LangGraph, FastAPI) | Ex-Founder | 50% latency reduction

Yashwanth Sai Sandeep Kalagatla  • Peerlist

Engineered a visually striking, high-performance digital presence for 64Framez, a creative media agency. This foundational client project was crucial in establishing Webortex's reputation for delivering polished, premium web solutions. ->Advanced Frontend Engineering: Implemented sophisticated front-end optimizations, including lazy loading for all media assets and on-scroll animations using the Intersection Observer API. This resulted in a 70% improvement in initial page load speed and a fluid, engaging user experience. -> Brand-Centric UX/UI: Architected a bespoke user interface with Tailwind CSS that captures the client's creative and modern brand identity, ensuring a seamless and responsive experience across all devices. -> SEO & Performance: Executed a comprehensive technical SEO strategy, leveraging the Server-Side Rendering (SSR) capabilities of Next.js to ensure maximum search engine visibility and achieve a Lighthouse performance score of 95+.

64Framez - High-Performance Creative Agency P

A platform to automate mark sheet generation and data handling in colleges. Key Contributions: ->Excel Automation: Designed a system to auto-generate student mark sheets with calculated grades. ->Database Integration: Oracle-based database to store and fetch student records. ->User Access: Secure teacher logins with data export/download capabilities. ->Tech Stack: Excel VBA, Oracle Database, React.js.

College Data Automation System(Collaborated)

👨💼 Developed a web-based Employee Management System to streamline HR operations. 💡 Key Contributions: ✅ CRUD Operations: Built a system to add, update, delete, and manage employees with a user-friendly UI. ✅ Authentication & Authorization: Implemented role-based access control (RBAC) for security. ✅ Data Handling: Used MongoDB & RESTful APIs for efficient data management. ✅ Tech Stack: Node.js, Express.js, React.js, MongoDB, JWT Authentication.

Employee Management System (EMS)

Career platform that uses RAG to give students AI-driven, actionable feedback on their project portfolios — not generic tips, but feedback grounded in their actual work. • Engineered a portfolio analysis engine using LlamaIndex and OpenAI API to retrieve and synthesize project-specific insights through RAG. • Built a MERN backend with JWT-based RBAC designed to support thousands of concurrent users with secure, isolated project data per user. • Launched in private beta with 50+ active users; iterating toward V1 public launch based on structured user feedback.

JobBridge — AI-Powered Career Platform

A full-stack platform I'm building from the ground up to bridge the gap between education and employment, leveraging AI to help students showcase their work and find opportunities. -> AI-Powered Guidance: Engineered a portfolio analysis feature using the OpenAI API and LlamaIndex for RAG, providing students with actionable, AI-driven feedback on their projects. -> Scalable Architecture: Architected a secure and scalable MERN stack backend with JWT-based authentication to support thousands of users and their project data. -> Current Status: Currently in private beta with 50+ active users, gathering feedback for a V1 public launch.

JobBridge - AI-Powered Career Platform

->Solid — add visualization type. ->Developed a stock analysis dashboard using external APIs to display real-time data. ->Implemented charts with libraries like Chart.js/D3.js and enabled technical indicators for investment decision-making.

Koinx Stock Analysis Dashboard

A reusable, high-conversion landing page template designed for personal brands and influencers, enabling them to launch a professional, aesthetically pleasing web presence in hours, not weeks. -> Reusable Architecture: Developed with a modular, component-based architecture in React & Tailwind CSS, allowing for rapid customization and deployment for multiple B2B clients. -> Premium UX/UI: Integrated Framer Motion for fluid, sophisticated animations that create a premium user experience and boost engagement metrics. ->Business Application: This template became a scalable product offering for Webortex, significantly reducing development time for a key client segment.

Productized Influencer Landing Page Template

Full-stack tool for storing, versioning, and A/B-testing LLM prompts built for personal use and actively used to iterate on retrieval templates for RAGLens AI. • Applied route-level code-splitting and dynamic imports across the Next.js app, Lighthouse performance score on the main dashboard improved from 61 to 94 on simulated 4G. • Evaluated Vercel Edge Functions vs Node.js serverless for prompt execution Edge cold-start measured at ~180ms vs ~1.1s for Node.js equivalent. Chose Edge for all latency-sensitive routes. • Supabase row-level security enforces per-user data isolation; all auth flows covered by integration tests.

PromptCraft Fusion — AI Prompt Engineering Pl

Production-grade multi-agent retrieval system that answers natural-language questions over a corpus of 8,000+ unstructured documents research papers and blog posts. • Built the full pipeline end-to-end: document ingestion → chunk embedding (OpenAI text-embedding-3-small) → pgvector similarity search → LLM synthesis → response with source citations. Full round-trip under 1.8s on local hardware. • Implemented Redis query caching measured 73% cache-hit rate on a 200-query test set, cutting average latency from 1.8s to 0.4s for repeated queries. • Built a custom semantic re-ranking layer and a 50-pair evaluation harness; improved retrieval precision@3 from 0.64 to 0.81, a 26.5% relative gain.

RAGLens AI — Multi-Agent RAG System

->Conceived, architected, and deployed a functional MVP for the "Researcho" academic journal platform for a client (Inside Journal) in a single, focused 2-hour sprint. -> Hyper-Efficient Execution: Leveraged a custom-built library of reusable components and a streamlined Next.js architecture to take the project from concept to a live, deployed URL in under two hours. -> Core Functionality Delivered: The initial MVP included a custom UI with a responsive design, a functional blog/article management system powered by a PHP/REST API backend, and foundational SEO optimizations. -> Immediate Business Impact: This rapid deployment allowed the client to secure immediate stakeholder buy-in and begin user testing weeks ahead of schedule, drastically accelerating their product development roadmap.

Rapid MVP Launch: "Researcho" Journal Platfor

->Built a full-stack restaurant management app enabling customers to browse menus, place orders, and track delivery status. ->Integrated role-based access for admins and chefs. ->Used MongoDB, Express.js, React.js, and Node.js to implement secure and scalable REST APIs.

Restaurant Management Application (MERN)

The primary digital presence and main sales engine for my company, Webortex, designed and engineered with a laser focus on technical SEO and inbound lead generation. -> Proven Business Impact: The site's SEO strategy is directly responsible for a >200% increase in our inbound client acquisition, serving as our primary case study and lead generator. -> Performance-Driven Tech: Built with Next.js for optimal Server-Side Rendering (SSR), achieving near-perfect Lighthouse scores and top-tier Core Web Vitals to maximize search engine ranking. -> Content & CMS: Integrated a headless CMS to allow for rapid content updates, which has been critical in securing projects with major clients