Jhapendra Kandel — Blog https://jhapendrakandel.com.np/blog/ 2026-09-06T00:00:00.000Z Jhapendra Kandel Building an AI-Powered Threat Intelligence Pipeline https://jhapendrakandel.com.np/blog/posts/ai-threat-intelligence-pipeline/ 2026-09-06T00:00:00.000Z Collect, normalize, analyze, enrich, summarize. Turning the threat-intel firehose into a short digest, with AI as one stage. Hello from the new blog https://jhapendrakandel.com.np/blog/posts/hello-from-the-new-blog/ 2026-09-03T00:00:00.000Z A short note on why this blog now lives on my own site, and how it works. Designing a Cybersecurity AI Agent Pipeline https://jhapendrakandel.com.np/blog/posts/cybersecurity-ai-agent-pipeline/ 2026-09-01T00:00:00.000Z Narrow specialized agents, evidence passed between stages, and a fact-check stage that assumes every finding is wrong. LLM Hallucinations in Cybersecurity: Why Verification Matters https://jhapendrakandel.com.np/blog/posts/llm-hallucinations-in-cybersecurity/ 2026-08-25T00:00:00.000Z Fabricated CVEs, plausible-but-wrong details, and made-up citations. Why every model claim in security work needs verifying. Using AI Agents for Cybersecurity: Where They Help and Where They Fail https://jhapendrakandel.com.np/blog/posts/ai-agents-for-cybersecurity/ 2026-08-12T00:00:00.000Z Where AI agents genuinely help in security work, where they fail with confidence, and how to use them without getting burned. Training a Nepali language model at home https://jhapendrakandel.com.np/blog/posts/nepali-llm-at-home/ 2026-08-05T00:00:00.000Z Notes on fine-tuning and serving a Nepali LLM on a single AMD AI Pro 32GB GPU in my homelab. Building a Security Research Assistant with Local LLMs https://jhapendrakandel.com.np/blog/posts/security-research-assistant-local-llms/ 2026-08-01T00:00:00.000Z A private security research helper on local models: RAG for knowledge, tool calling for facts, and keeping output honest. What Happens When an Internet-Facing Server Gets Attacked? https://jhapendrakandel.com.np/blog/posts/internet-facing-server-attacked/ 2026-07-23T00:00:00.000Z The constant background noise, what the attack surface really is, why logs must leave the box, and a short IR plan. Why Production Engineering Is Different from Writing Code https://jhapendrakandel.com.np/blog/posts/production-engineering-vs-code/ 2026-07-14T00:00:00.000Z Writing code that works is the small part. Reliability, monitoring, backups, security and deployment are the rest. From Localhost to Production: Problems Nobody Mentions in Tutorials https://jhapendrakandel.com.np/blog/posts/localhost-to-production/ 2026-07-01T00:00:00.000Z DNS, service networking, permissions, containers and env config. The things that break the first few times you ship for real. Why Network Isolation Matters in a Cybersecurity Lab https://jhapendrakandel.com.np/blog/posts/network-isolation-security-lab/ 2026-06-17T00:00:00.000Z VLANs, default-deny between segments, and small-scale zero trust. Why isolation keeps a lab from becoming an incident. How I Think About Backups After Designing My Own Infrastructure https://jhapendrakandel.com.np/blog/posts/how-i-think-about-backups/ 2026-06-03T00:00:00.000Z Tiers not copies, the target pulls, keep history, test the restore. The backup model I use after building my own infra. What I Learned Designing a Multi-Layer Server Architecture https://jhapendrakandel.com.np/blog/posts/multi-layer-server-architecture/ 2026-05-19T00:00:00.000Z A VPS, a NAS, a backup box and some tunnels, each with one job. Why the separation happened and what it costs. Cloudflare Tunnels as Part of a Self-Hosted Infrastructure https://jhapendrakandel.com.np/blog/posts/cloudflare-tunnels-self-hosted/ 2026-05-06T00:00:00.000Z Exposing self-hosted services without opening ports: what tunnels buy you, how I use them, and where they fall short. Can Multiple AI Agents Work in Parallel? Understanding Parallel Agent Architectures https://jhapendrakandel.com.np/blog/posts/parallel-ai-agents/ 2026-04-21T00:00:00.000Z When running AI agents in parallel actually helps, the architecture that works, and the judging step everyone skips. Why Nepali Is Still a Hard Language for AI Models https://jhapendrakandel.com.np/blog/posts/why-nepali-is-hard-for-ai/ 2026-04-02T00:00:00.000Z Thin data, tokenizers built for English, and manual evaluation. The concrete reasons Nepali still lags, and what helps. Running Large Language Models Locally: What VRAM Actually Means https://jhapendrakandel.com.np/blog/posts/running-llms-locally-vram/ 2026-03-18T00:00:00.000Z Parameters are not gigabytes. Where VRAM really goes, why quantization works, and why a 70B model does not need 70GB.