---
title: "AI Developer (@lycore) | Hashnode"
description: "build production-ready AI integrations, mobile apps, and web platforms for businesses that want real results."
url: "https://hashnode.com/@lycore"
---

@lycore

# AI Developer

@lycore · Joined April 2026

build production-ready AI integrations, mobile apps, and web platforms for businesses that want real results.

### [Hide, Hold or Allow: AI Moderation Helpers in Django (Part 8)](https://aidevpro.hashnode.dev/ai-content-moderation-django-appeals?utm_source=hashnode&utm_medium=feed)

Regex filters that only hide what's never innocent, an LLM verdict as a validated Pydantic output with a policy clause and evidence quote, thresholds that fail to a human, and an appeals queue whose overturns become eval cases.

18 min read · [aidevpro.hashnode.dev](https://aidevpro.hashnode.dev/ai-content-moderation-django-appeals?utm_source=hashnode&utm_medium=feed)

[1h ago](https://hashnode.com/posts/ai-content-moderation-django-appeals/6aff746a2e356661a1e97b9f)

### [Java4Shell (CVE-2026-114938): Patch Your JVMs or Flip One Flag Today](https://aidevpro.hashnode.dev/java4shell-cve-2026-114938-openjdk-patch-guide?utm_source=hashnode&utm_medium=feed)

Java4Shell hits OpenJDK 17.0.6+, 21, 25, 27 on AVX-512 x86-64. Find every JVM, patch, or set -XX:-UseBASE64Intrinsics and verify it.

20 min read · [aidevpro.hashnode.dev](https://aidevpro.hashnode.dev/java4shell-cve-2026-114938-openjdk-patch-guide?utm_source=hashnode&utm_medium=feed)

[1d ago](https://hashnode.com/posts/java4shell-cve-2026-114938-openjdk-patch-guide/6afd7e98c1fcba50755a4eda)

### [Valid JSON, Wrong Refund: Structured LLM Outputs in Django (Part 7)](https://aidevpro.hashnode.dev/structured-llm-outputs-pydantic-django?utm_source=hashnode&utm_medium=feed)

Constrained decoding guarantees shape, not truth: four validation layers, fact checks through Pydantic validation context, one model for Anthropic and OpenAI schemas, and a retry policy that knows when to stop.

18 min read · [aidevpro.hashnode.dev](https://aidevpro.hashnode.dev/structured-llm-outputs-pydantic-django?utm_source=hashnode&utm_medium=feed)

[2](https://hashnode.com/posts/structured-llm-outputs-pydantic-django/6afcd459c7d39edf63adb6b0) [2d ago](https://hashnode.com/posts/structured-llm-outputs-pydantic-django/6afcd459c7d39edf63adb6b0)

### [Don't Let the Meeting Bot File the Ticket: Building Around Anthropic's Claude Meetings Beta](https://aidevpro.hashnode.dev/claude-meetings-review-before-you-file-the-ticket?utm_source=hashnode&utm_medium=feed)

Claude Meetings joins Google Meet as a visible participant. How engineering teams should handle consent, connected tools, and ticket review.

18 min read · [aidevpro.hashnode.dev](https://aidevpro.hashnode.dev/claude-meetings-review-before-you-file-the-ticket?utm_source=hashnode&utm_medium=feed)

[6d ago](https://hashnode.com/posts/claude-meetings-review-before-you-file-the-ticket/6af69705337f1f4696e0d9ec)

### [Don't Put the Prompt on the Span: Tracing Django AI Agents with OpenTelemetry (Part 6)](https://aidevpro.hashnode.dev/opentelemetry-tracing-ai-agents-django?utm_source=hashnode&utm_medium=feed)

Trace Django AI agents with OpenTelemetry GenAI conventions: PII-safe spans, prompt refs, tail sampling on cost, and debugging a bad answer.

15 min read · [aidevpro.hashnode.dev](https://aidevpro.hashnode.dev/opentelemetry-tracing-ai-agents-django?utm_source=hashnode&utm_medium=feed)

[Nov 10](https://hashnode.com/posts/opentelemetry-tracing-ai-agents-django/6af396b662b23cd0618a9d4e)
