---
title: "Precious Obinna (@Prehqsy) | Hashnode"
description: "Developer. Read the latest articles by Precious Obinna on Hashnode."
url: "https://hashnode.com/@Prehqsy"
---

@Prehqsy

# Precious Obinna

@Prehqsy · Lagos Nigeria · Joined February 2024

Developer

### [How I Used Context Engineering to Improve Retrieval](https://precious-o.hashnode.dev/how-i-used-context-engineering-to-improve-retrieval?utm_source=hashnode&utm_medium=feed)

How I used context engineering to enrich documents and queries so semantic search actually finds the right results. With Python code and examples.

8 min read · [precious-o.hashnode.dev](https://precious-o.hashnode.dev/how-i-used-context-engineering-to-improve-retrieval?utm_source=hashnode&utm_medium=feed)

[Jun 12](https://hashnode.com/posts/how-i-used-context-engineering-to-improve-retrieval/6a2c16288e129d96e4f58029)

### [How I cached intention, not queries](https://precious-o.hashnode.dev/how-i-cached-intention-not-queries?utm_source=hashnode&utm_medium=feed)

Cut LLM token usage 80-90% and latency from 40s to 10s by caching user intent with semantic embeddings and Qdrant, not exact queries

3 min read · [precious-o.hashnode.dev](https://precious-o.hashnode.dev/how-i-cached-intention-not-queries?utm_source=hashnode&utm_medium=feed)

[Jun 3](https://hashnode.com/posts/how-i-cached-intention-not-queries/6a1f886d1d44cf54bad48e57)

### [Threads vs Processes in Task Queues: The Reliability Tradeoff](https://precious-o.hashnode.dev/threads-vs-processes-in-task-queues-the-reliability-tradeoff?utm_source=hashnode&utm_medium=feed)

Why thread-based task queues fail in production — fake timeouts, resource leaks — and how processes fix it with real isolation and safe termination.

3 min read · [precious-o.hashnode.dev](https://precious-o.hashnode.dev/threads-vs-processes-in-task-queues-the-reliability-tradeoff?utm_source=hashnode&utm_medium=feed)

[May 15](https://hashnode.com/posts/threads-vs-processes-in-task-queues-the-reliability-tradeoff/6a0673379c794cb653cbc3f1)

### [The Reflection Pattern in AI: Teaching Models to Think About Their Thinking](https://precious-o.hashnode.dev/the-reflection-pattern-in-ai-teaching-models-to-think-about-their-thinking?utm_source=hashnode&utm_medium=feed)

Learn the Reflection Pattern: teach AI to self-critique and refine outputs for higher accuracy, fewer errors, and stronger reasoning.

4 min read · [precious-o.hashnode.dev](https://precious-o.hashnode.dev/the-reflection-pattern-in-ai-teaching-models-to-think-about-their-thinking?utm_source=hashnode&utm_medium=feed)

[Feb 10](https://hashnode.com/posts/the-reflection-pattern-in-ai-teaching-models-to-think-about-their-thinking/698af163145323b23a103e4e)

### [Retrieval Is Not Resolution: Building a Hallucination-Resistant RAG System with LLMs and SQL Server](https://precious-o.hashnode.dev/retrieval-is-not-resolution-building-a-hallucination-resistant-rag-system-with-llms-and-sql-server?utm_source=hashnode&utm_medium=feed)

Step-by-step breakdown of a production-style RAG system for gadget price lookup. Qdrant retrieves candidates, a deterministic resolver prevents wrong match

6 min read · [precious-o.hashnode.dev](https://precious-o.hashnode.dev/retrieval-is-not-resolution-building-a-hallucination-resistant-rag-system-with-llms-and-sql-server?utm_source=hashnode&utm_medium=feed)

[Dec 31, 2025](https://hashnode.com/posts/retrieval-is-not-resolution-building-a-hallucination-resistant-rag-system-with-llms-and-sql-server/695525755c22d193c247378a)
