# Pradeep Bhandari — Senior Engineering Manager & Full-Stack Architect > Personal site of Pradeep Bhandari, Senior Engineering Manager and Full-Stack Architect. Writing on vector search, > data pipelines and applied AI, plus a suite of free, privacy-first developer tools. Site: https://pradeepbhandari.com Last updated: 2026-09-04 --- ## About Pradeep Bhandari is a Senior Engineering Manager and Full-Stack Architect with a decade of experience turning complex technical challenges into resilient, production-grade applications and reliable data systems. Core expertise spans PHP / Laravel, Python, vector search (pgvector, Vespa.ai), event-driven pipelines (Kafka, Airflow), and JavaScript including Three.js / WebGL. --- ## Pages - [Home](https://pradeepbhandari.com/): Portfolio homepage. About section, selected projects, Labs preview, Lumina preview, contact. - [Labs](https://pradeepbhandari.com/labs): Suite of 8 free, privacy-first, client-side developer utilities. All processing runs in the browser — no data is sent to any server. - [Prism](https://pradeepbhandari.com/prism): AI creative studio. Turns a prompt into ad creatives for Instagram, Stories, and landscape formats, then lets you edit the text layers directly in the browser. - [Lumina](https://pradeepbhandari.com/lumina): Free web-based mockup tool. Paste any URL, get a 3D perspective website screenshot with custom gradient backgrounds and browser frames. Export at 4K. - [Blog](https://pradeepbhandari.com/blog): Technical articles on software engineering, creative development, Three.js, Laravel, and building developer tools. - [Sitemap](https://pradeepbhandari.com/sitemap.xml) - [RSS Feed](https://pradeepbhandari.com/feed.xml) --- ## Blog Articles - [The Local Embedding Edge: Replacing OpenAI with Self-Hosted SLMs in Laravel RAG Pipelines](https://pradeepbhandari.com/blog/self-hosted-embedding-model-rag-laravel-ollama): Transition from OpenAI to self-hosted embedding models in Laravel 13. This guide covers Ollama integration, pgvector dimension management, and high-throughput Python sidecars. - [The Semantic Firewall: Hardening Laravel 13 Agents Against Prompt Injection](https://pradeepbhandari.com/blog/laravel-13-llm-prompt-injection-protection-guide): Build a resilient Laravel 13 agent by implementing a Semantic Firewall with Llama Guard to filter untrusted vector data and prevent indirect prompt injection attacks. - [The Freshness Problem: Real-Time Vector Sync with Debezium and Kafka](https://pradeepbhandari.com/blog/change-data-capture-for-vector-search-kafka): Move RAG synchronization to the infrastructure level. Learn how Change Data Capture for Vector Search using Debezium and Kafka solves the dual-write problem and Update Storms. - [Hard Isolation for Vectors: A Hands-On pgvector Laravel 13 Setup Guide](https://pradeepbhandari.com/blog/pgvector-laravel-setup-guide-multi-tenant-rls): Secure your multi-tenant RAG systems by implementing PostgreSQL Row Level Security with pgvector and Laravel 13. Learn to enforce hard isolation at the database level. - [Beyond Semantic Similarity: Implementing GraphRAG with pgvector and Laravel](https://pradeepbhandari.com/blog/graphrag-laravel-implementation-pgvector): Learn how to implement GraphRAG in Laravel using pgvector and Recursive CTEs. Move beyond flat similarity search by extracting entity triples for multi-hop AI reasoning. - [Graduating Your RAG Ingestion: When Laravel Queues Hit the Ceiling and Airflow Takes Over](https://pradeepbhandari.com/blog/apache-airflow-rag-pipeline-orchestration-laravel): Moving RAG ingestion from Laravel Horizon to Apache Airflow ensures resilience for complex vector pipelines. Learn to build a 'Signal-and-Poll' pattern for robust AI data orchestra... - [The Silent Search Killer: Using dbt Tests to Guard Your Vector Ingestion Pipeline](https://pradeepbhandari.com/blog/dbt-for-vector-data-quality-rag-pipelines): Stop silent search degradation by implementing dbt macros to validate vector dimensions, L2 norms, and semantic drift across your ingestion pipelines. - [The Privacy Moat: Implementing Hard Multi-Tenancy in pgvector with PostgreSQL RLS](https://pradeepbhandari.com/blog/pgvector-multi-tenancy-postgresql-rls-guide): Secure your RAG architecture by moving from application-side metadata filtering to database-level isolation using pgvector multi-tenancy, PostgreSQL RLS, and HNSW optimization. - [Livewire v4 + Reverb: Streaming Asynchronous AI Agent Responses Word-by-Word](https://pradeepbhandari.com/blog/livewire-v4-reverb-streaming-laravel-ai-sdk): An extensive technical layout tutorial constructing an interactive, token-streaming interface using Laravel 13. We walk through extracting real-time generative updates from the new... - [Hybrid Retrieval 2.0: Tuning HNSW Parameters for pgvector at Scale](https://pradeepbhandari.com/blog/pgvector-hnsw-performance-tuning-scaling-guide): Optimize pgvector HNSW parameters for million-row vector databases. This guide covers m, ef_construction, and memory management for production-grade RAG. - [Hallucination-Proofing Your Agent: Building a RAG Evaluation Framework in PHP](https://pradeepbhandari.com/blog/rag-evaluation-framework-php-laravel): Eliminate AI hallucinations by building a native RAG evaluation framework in Laravel 13 using Pest, LLM-as-a-judge, and quantitative metrics to verify agent faithfulness. - [Model Context Protocol in Laravel: Building MCP Servers with the AI SDK](https://pradeepbhandari.com/blog/model-context-protocol-laravel-building-mcp-servers-ai-sdk): Standardize your AI integrations by building MCP servers in Laravel 13. Learn to map tools to resources and implement SSE transports for high-scale Kafka and vector data pipelines. - [Architecting for the Bill: Building an LLM Rate Limiter and Cost Tracker in Laravel](https://pradeepbhandari.com/blog/llm-cost-management-laravel-rate-limiter): Learn how to build a robust LLM cost management system in Laravel 13 using middleware to track tokens, manage multi-tenant budgets, and prevent runaway costs. - [Event-Driven Ingestion: Scaling Search Pipelines with Kafka and dbt](https://pradeepbhandari.com/blog/event-driven-search-ingestion-kafka-dbt-pipeline): An enterprise data architecture tutorial focused on building a high-throughput search ingestion pipeline. We demonstrate how to break up transactional database dependencies by usin... - [Two-Stage Retrieval: Cross-Encoder Reranking with Laravel 13 AI SDK](https://pradeepbhandari.com/blog/laravel-13-ai-sdk-cross-encoder-reranking): An deep-dive architectural tutorial demonstrating how to combine first-stage vector recall via pgvector with second-phase precision sorting using the Laravel 13 AI SDK Reranking en... - [Location-Aware RAG: Combining Uber’s H3 Grids with Laravel 13 Vector Search](https://pradeepbhandari.com/blog/laravel-13-location-aware-rag-uber-h3-pgvector): An extensive system architecture tutorial analyzing how to construct location-aware Retrieval-Augmented Generation engines within the Laravel 13 ecosystem. We detail how to impleme... - [Beyond Wrappers: Building Autonomous Agents with Laravel 13 AI SDK](https://pradeepbhandari.com/blog/laravel-13-ai-sdk-agent-architecture-guide): An extensive end-to-end guide exploring the newly stabilized Laravel 13 AI SDK. We walk through setting up dedicated Artisan Agent classes, enabling native database conversation pe... - [Asynchronous Embedding Ingestion: Scaling pgvector in Laravel 13 with Queue Batches](https://pradeepbhandari.com/blog/laravel-13-pgvector-asynchronous-queue-batches): An infrastructure guide focused on building a high-throughput, decoupled data ingestion pipeline using Laravel 13. We show how to safely handle high-volume text chunking, eliminate... - [Real-Time Ingestion: Syncing Data to Vespa Without Lag](https://pradeepbhandari.com/blog/real-time-ingestion-syncing-data-to-vespa-without-lag): An architectural guide to implementing real-time data synchronization with Vespa. We break down the dangers of synchronous vector generation and provide an event-driven blueprint t... - [Building a Real-Time Search Dashboard with Laravel 13 & Reverb](https://pradeepbhandari.com/blog/building-a-real-time-search-dashboard-with-laravel-13-reverb): A hands-on, end-to-end tutorial detailing how to configure a real-time operational dashboard using Laravel 13 and Laravel Reverb. We bypass legacy database polling architectures by... - [Vespa Multi-Stage Ranking: Merging BM25 and Vector Search](https://pradeepbhandari.com/blog/vespa-multi-stage-ranking-merging-bm25-and-vector-search): An in-depth technical tutorial detailing how to craft hybrid, multi-stage rank profiles inside Vespa.ai. It breaks down the mechanical logic of first-phase (lexical) and second-pha... - [Vespa Streaming Mode: The Architect’s Guide to Cost-Efficient Private Search](https://pradeepbhandari.com/blog/vespa-streaming-mode-the-architects-guide-to-cost-efficient-private-search): An architectural exploration of Vespa Streaming Mode. We discuss the cost-efficiency of "index-less" search for private, multi-tenant data and provide a framework for architects to... - [Building a Semantic Newsletter Engine with Laravel 13 AI SDK](https://pradeepbhandari.com/blog/building-a-semantic-newsletter-engine-with-laravel-13-ai-sdk): A comprehensive guide to building an AI-powered newsletter engine. We demonstrate how to leverage Laravel 13's native vector search and the first-party AI SDK to create a curation... - [The 'Edit' Button for AI: Implementing Multi-Step Approval Workflows in Laravel 13 RAG](https://pradeepbhandari.com/blog/laravel-13-rag-human-in-the-loop-workflow): Autonomous AI is great until it hallucinates a refund policy. Learn how to use the Laravel 13 AI SDK and Livewire v4 to build a "Human-in-the-Loop" search engine. We'll show you ho... - [Building AI-Native Apps: Vector Search in Laravel 13 & PostgreSQL](https://pradeepbhandari.com/blog/native-vector-search-laravel-13-eloquent): This guide explores how Laravel 13 has revolutionized search by introducing native vector similarity methods. We cover everything from the basic whereVectorSimilarTo syntax to the... - [FastAPI vs. Prism vs. WireMock: The 2026 Developer’s Guide to Mastering API Mocking](https://pradeepbhandari.com/blog/fastapi-vs-prism-vs-wiremock-2026-guide): Stop stalling your development cycle. Discover why FastAPI, Prism, and WireMock are the three pillars of 2026 API mocking—perfect for testing AI agents, ensuring contract certainty... - [Beyond the Single Prompt: Orchestrating Multi-Step AI Workflows in Laravel 13](https://pradeepbhandari.com/blog/beyond-the-single-prompt-orchestrating-multi-step-ai-workflows-in-laravel-13): One AI agent is a chatbot; three AI agents are a department. In this tutorial, we move beyond basic API calls to build a sophisticated "Orchestrator-Worker" system. See how to use... - [Step-by-Step: Building a Zero-Downtime Data Migration Tool with Laravel Zero](https://pradeepbhandari.com/blog/laravel-zero-zero-downtime-migration-tool): Stop crashing your production DB. Standard migrations are fine for small apps, but when you hit millions of rows, you need a strategy. Learn how to build a custom CLI migration too... - [Native Vector Search in Eloquent: Semantic Discovery the Easy Way](https://pradeepbhandari.com/blog/laravel-native-vector-search-eloquent-pgvector): Why manage a massive external search cluster when your favorite framework can handle it natively? In 2026, Laravel and pgvector have finally merged to give us "Semantic Search" wit... - [Beyond the Dashboard: Engineering a High-Throughput GMB Data Pipeline](https://pradeepbhandari.com/blog/engineering-gmb-api-data-pipeline-at-scale): Managing a few local listings is easy; orchestrating thousands across a national network is a distributed systems challenge. Here’s how I build resilient, reactive GMB infrastructu... - [Beyond the Radius: Why I’m Switching Dealer Networks to Hexagonal Grids](https://pradeepbhandari.com/blog/hexagonal-grids-vs-point-in-polygon-geospatial-engineering): Traditional "Point-in-Polygon" searches are failing modern dealer networks. As I transition into Data Engineering, I’ve found that administrative boundaries are too "noisy" for rea... - [Vespa.ai vs Elasticsearch: Which is Best for Real-Time Vector Search?](https://pradeepbhandari.com/blog/vespa-ai-vs-elasticsearch-real-time-vector-search): A deep dive into Vespa.ai vs Elasticsearch for 2026. Learn why Vespa's native tensor support and AI-driven query conversion win for real-time search intelligence. - [Next.js vs. Laravel: Which Framework Should You Choose in 2026?](https://pradeepbhandari.com/blog/nextjs-vs-laravel-which-framework-should-you-choose-in-2026): In 2026, the "PHP is dead" meme is long gone, and Next.js has matured past its experimental App Router phase. --- ## Tools (Labs) All eight run entirely client-side in the browser — no upload, no server, no tracking. - [Open Graph & Meta Studio](https://pradeepbhandari.com/labs/open-graph): Previews and generates Open Graph, Twitter Card, and meta tags for any URL. - [Code Screenshot Generator](https://pradeepbhandari.com/labs/code-screenshot): Turns source code into shareable images with syntax themes, window frames, and PNG export. - [Image Optimizer](https://pradeepbhandari.com/labs#tiny-compress): Lossless browser-side PNG / JPG / WebP compression with batch ZIP download. Max 10 images. - [SVG Nano](https://pradeepbhandari.com/labs#svg-nano): SVG minifier — strips comments, redundant attributes, and whitespace. - [JSON to TypeScript](https://pradeepbhandari.com/labs#schema-weaver): Converts a JSON object or array into TypeScript interface definitions. - [Fluid Typography](https://pradeepbhandari.com/labs#fluid-type): Generates a responsive CSS `clamp()` font-size from min/max px values. - [cURL Transformer](https://pradeepbhandari.com/labs#curl-transformer): Converts cURL commands to Fetch API, Axios, Python Requests, Go, or Laravel HTTP. - [Crontab Generator](https://pradeepbhandari.com/labs#cron-gen): Visual cron expression builder with human-readable output and common presets. --- ## Projects - [Indian Pincodes](https://pradeepbhandari.com/#work): Laravel REST API for searching Indian postal codes, backed by Redis and PostgreSQL. - [Grid Search](https://pradeepbhandari.com/#work): Geospatial visualisation tool built with React, Three.js, and the Google Maps API. - [Fetch Reviews](https://pradeepbhandari.com/#work): Python web scraper performing NLP sentiment analysis on product reviews. --- ## Contact & Socials - [Email](mailto:me@pradeepbhandari.com) - [GitHub](https://github.com/swankey) - [LinkedIn](https://linkedin.com/in/bhands) --- ## AI Usage Policy All public content on this site (blog posts, tool descriptions, portfolio) is available for AI training and indexing. The admin area (/admin/) is excluded. See /robots.txt for per-bot crawl rules.