---
title: "Dhananjay Lakkawar (@lakkawardhananjay) | Hashnode"
description: "Cloud Architect & AI Engineer. Building scalable systems on cloud. Read the latest articles by Dhananjay Lakkawar on Hashnode."
url: "https://hashnode.com/@lakkawardhananjay"
---

@lakkawardhananjay

# Dhananjay Lakkawar

@lakkawardhananjay · Joined November 2024

Cloud Architect & AI Engineer. Building scalable systems on cloud.

### [The 15-Millisecond AI: Building "Pre-Cognitive" Edge Caching on AWS](https://genaiguru.hashnode.dev/the-15-millisecond-ai-building-pre-cognitive-edge-caching-on-aws?utm_source=hashnode&utm_medium=feed)

Stop making users wait for AI to type. Learn how to build a "pre-cognitive" AI architecture using AWS Step Functions, Bedrock, and CloudFront Lambda@

5 min read · [genaiguru.hashnode.dev](https://genaiguru.hashnode.dev/the-15-millisecond-ai-building-pre-cognitive-edge-caching-on-aws?utm_source=hashnode&utm_medium=feed)

[Mar 29](https://hashnode.com/posts/the-15-millisecond-ai-building-pre-cognitive-edge-caching-on-aws/69c97a385282e6ac9ce28a63)

### [The $50,000 Chat History Problem: Building Event-Driven AI Memory on AWS](https://genaiguru.hashnode.dev/the-50-000-chat-history-problem-building-event-driven-ai-memory-on-aws?utm_source=hashnode&utm_medium=feed)

Stop stuffing chat history into expensive LLM context windows. Learn how to architect a highly scalable, event-driven AI memory system using AWS Event

5 min read · [genaiguru.hashnode.dev](https://genaiguru.hashnode.dev/the-50-000-chat-history-problem-building-event-driven-ai-memory-on-aws?utm_source=hashnode&utm_medium=feed)

[Mar 27](https://hashnode.com/posts/the-50-000-chat-history-problem-building-event-driven-ai-memory-on-aws/69c6acddae4d2d0c2f448c5e)

### [Routing LLM Traffic on AWS: How to Build a Cost-Optimized Multi-Model API Router](https://genaiguru.hashnode.dev/cost-optimized-multi-model-api-router?utm_source=hashnode&utm_medium=feed)

Stop overpaying for high-end LLMs. Learn how to architect a serverless API router on AWS using Bedrock, Lambda, and API Gateway to cut AI inference .

5 min read · [genaiguru.hashnode.dev](https://genaiguru.hashnode.dev/cost-optimized-multi-model-api-router?utm_source=hashnode&utm_medium=feed)

[Mar 18](https://hashnode.com/posts/cost-optimized-multi-model-api-router/69ba8228c22d3eeb8a734a8e)

### [Stop Overpaying for VectorDBs: Architecting Serverless RAG on AWS](https://genaiguru.hashnode.dev/stop-overpaying-for-vectordbs-architecting-serverless-rag-on-aws?utm_source=hashnode&utm_medium=feed)

Build a production-ready Retrieval-Augmented Generation pipeline using Amazon Bedrock, AWS Lambda, and OpenSearch Serverless while eliminating idle infrastructure costs.

5 min read · [genaiguru.hashnode.dev](https://genaiguru.hashnode.dev/stop-overpaying-for-vectordbs-architecting-serverless-rag-on-aws?utm_source=hashnode&utm_medium=feed)

[Mar 14](https://hashnode.com/posts/stop-overpaying-for-vectordbs-architecting-serverless-rag-on-aws/69b508dc210c74252fc77320)

### [The GenAI Paradox: Why “Buying” Managed Infrastructure Is the Only Way to “Build” Effectively](https://genaiguru.hashnode.dev/the-genai-paradox-why-buying-managed-infrastructure-is-the-only-way-to-build-effectively?utm_source=hashnode&utm_medium=feed)

A CTO’s perspective on the real architectural trade-offs between self-hosted LLM stacks and serverless inference chains.

4 min read · [genaiguru.hashnode.dev](https://genaiguru.hashnode.dev/the-genai-paradox-why-buying-managed-infrastructure-is-the-only-way-to-build-effectively?utm_source=hashnode&utm_medium=feed)

[Jan 23](https://hashnode.com/posts/the-genai-paradox-why-buying-managed-infrastructure-is-the-only-way-to-build-effectively/6973c324df45f8d000797a07)