--- title: "Ajitabh Singh (@ajitabh) | Hashnode" description: "Real-world engineering insights on AI, systems, and scalable design. Read the latest articles by Ajitabh Singh on Hashnode." url: "https://hashnode.com/@ajitabh" --- @ajitabh # Ajitabh Singh @ajitabh · New York, USA · Joined March 2026 Real-world engineering insights on AI, systems, and scalable design ## Comments [Great point. Chunking strategies should really be evaluated against evolving query patterns, not just a static benchmark. Stability under changing retrieval needs is an important part of production readiness.](https://hashnode.com/posts/why-chunking-strategy-can-make-or-break-your-rag-pipeline/6a7ab6663ddc972e303c1f03/comment/6a7c324236c4c02836a8f2f7#reply-6a7dfb0f9870a68064ac2564) Reply · Article · Aug 13 · [Why Chunking Strategy Can Make or Break Your RAG Pipeline](https://hashnode.com/posts/why-chunking-strategy-can-make-or-break-your-rag-pipeline/6a7ab6663ddc972e303c1f03) [Absolutely agree. I don't think chunk size should be tuned by eye. My recommendation is to align chunks with the document's natural structure first, then compare strategies using retrieval metrics like Recall@K, Precision@K, and MRR on a representative query set. The key point for me is that chunk size is a workload-dependent decision, not a universal number.](https://hashnode.com/posts/why-chunking-strategy-can-make-or-break-your-rag-pipeline/6a7ab6663ddc972e303c1f03/comment/6a7b551942d6c75df9575550#reply-6a7b8266b2f61ffc0969b0b6) Reply · Article · Aug 11 · [Why Chunking Strategy Can Make or Break Your RAG Pipeline](https://hashnode.com/posts/why-chunking-strategy-can-make-or-break-your-rag-pipeline/6a7ab6663ddc972e303c1f03) [Andrew Fair points, but I think these examples actually reinforce the main point of the article. I’m not arguing that hierarchical or overlap-based chunking is universally better. The article explicitly says each strategy has trade-offs and that the right choice depends on the document and retrieval workload. The hierarchical example is a valid implementation failure, but not necessarily a failure of hierarchical chunking itself. Likewise, I explicitly call out that overlap reduces boundary impact but doesn't eliminate semantic splits. I agree that retrieval should be evaluated independently before generation. Ultimately, the goal is simple: test different strategies on representative queries and choose the one that gives the best retrieval quality for the document type, while considering cost and complexity.](https://hashnode.com/posts/why-chunking-strategy-can-make-or-break-your-rag-pipeline/6a7ab6663ddc972e303c1f03/comment/6a7b6bb866cbedfec779fb73#reply-6a7b820f91e3db29be6c2d35) Reply · Article · Aug 11 · [Why Chunking Strategy Can Make or Break Your RAG Pipeline](https://hashnode.com/posts/why-chunking-strategy-can-make-or-break-your-rag-pipeline/6a7ab6663ddc972e303c1f03) [Spot on. Fast build is easy, maintainable systems are the hard part.](https://hashnode.com/posts/the-vibe-coding-trap-ai-architecture/69dd08e13000edd30fdb3991/comment/69dded4fdebb505138053b1b#reply-69fa5af1a386d7f121d8c29a) Reply · Article · May 5 · [The Vibe Coding Trap: Why AI-Driven Development Needs Strong Architecture](https://hashnode.com/posts/the-vibe-coding-trap-ai-architecture/69dd08e13000edd30fdb3991) [SSorry for the late response , this really resonates. I’ve seen the same thing. AI code is fast to build but can be hard to maintain later. Setting clear architecture first feels important. ADRs could help as a good guardrail, as long as they stay simple and are actually followed in practice.](https://hashnode.com/posts/the-vibe-coding-trap-ai-architecture/69dd08e13000edd30fdb3991/comment/69de51c0345b86c2e05175d6#reply-69fa5aa6a386d7f121d89e72) Reply · Article · May 5 · [The Vibe Coding Trap: Why AI-Driven Development Needs Strong Architecture](https://hashnode.com/posts/the-vibe-coding-trap-ai-architecture/69dd08e13000edd30fdb3991) [Yeah. The trap isn’t fast code, it’s outsourcing thinking to speed.](https://hashnode.com/posts/the-vibe-coding-trap-ai-architecture/69dd08e13000edd30fdb3991/comment/69df8bbc70e165e0d8a836cd#reply-69fa5a1fa386d7f121d85932) Reply · Article · May 5 · 1 · [The Vibe Coding Trap: Why AI-Driven Development Needs Strong Architecture](https://hashnode.com/posts/the-vibe-coding-trap-ai-architecture/69dd08e13000edd30fdb3991) [The “junior dev PR” framing is exactly it. AI gets you to day one fast, but maintainability still comes from design, not generation. Without that, you’re just deferring the cost.](https://hashnode.com/posts/the-vibe-coding-trap-ai-architecture/69dd08e13000edd30fdb3991/comment/69dfa32370e165e0d8bb9a63#reply-69fa59efa386d7f121d83d59) Reply · Article · May 5 · [The Vibe Coding Trap: Why AI-Driven Development Needs Strong Architecture](https://hashnode.com/posts/the-vibe-coding-trap-ai-architecture/69dd08e13000edd30fdb3991) [100%. AI scales output, not clarity.](https://hashnode.com/posts/the-vibe-coding-trap-ai-architecture/69dd08e13000edd30fdb3991/comment/69de0672c444e3272334711c#reply-69fa598da386d7f121d807d1) Reply · Article · May 5 · 1 · [The Vibe Coding Trap: Why AI-Driven Development Needs Strong Architecture](https://hashnode.com/posts/the-vibe-coding-trap-ai-architecture/69dd08e13000edd30fdb3991) [Glad you found it interesting! It’s a newer way to think about RAG that does not use vector embeddings. I’d love to hear your thoughts if you have used traditional RAG before. Thanks!](https://hashnode.com/posts/rag-vs-vectorless-rag/69c0ce24d9da55a9a5374a93/comment/69c1b06530a9b81e3aaac5f5#reply-69c29a3392029d915f326be7) Reply · Article · Mar 24 · 1 · [RAG vs Vectorless RAG: How AI Systems Retrieve Knowledge](https://hashnode.com/posts/rag-vs-vectorless-rag/69c0ce24d9da55a9a5374a93)