Nicheloom visits: 60,536 all-time · 7,913 in the last 7 days

The opportunity tracker for new startups.

Today: 895 launches · Niche of the day: browser-based video games · See what launched today →

Skald

open-source context layer API that runs in your VPC

Get picks like this daily. Today's launch numbers, the top launches, AI and tech news, and a daily opportunity pick, straight to your inbox.

Want to reach these readers? Sponsor an issue.

Hey HN! We built an open source context layer to power AI agents and apps even in air-gapped infra setups.TL;DR: Use our API (or one of the 6 SDKs) to push context into the Skald platform, and get semantic search and AI chat out of the box.We’ve seen companies spend months building a context layer system internally, only for it to have subpar performance and require active maintenance.Skald gives you the plumbing to get started really fast when building context-aware agents and AI apps (customers have gone to prod in a day with us) but is still highly extensible and configurable to fit individual needs.Our core is entirely MIT-licensed and you can even run it with a fully self-hosted stack (document parsing, chunking, vector search, LLM, etc) in your VPC. We think solutions in this space must be open source, and not just “open source but bring 6 API keys to get started”.Our customers have used Skald to quickly build features for their own users, as well as for internal tooling.The self-hosted MIT version is completely fully-featured, so much so that it’s what we run on our Cloud offering today. We do also have an enterprise license for larger customers that runs on their infra and has a few enterprise-focused features on top of the MIT version.Keen to hear your feedback and I’ll be around to answer questions!GitHub: https://github.com/skaldlabs/skald

HN · launched Dec 10, 2025 · 5 upvotes (top 90%) · 0 views

Niche: semantic search and rag tools · 93 launches, 0 in the last 7 days · Explore this niche →

Category and tags

  • Horizontal
  • Data infrastructure
  • Developer
  • AI-native
  • Commercial product

Visit website ↗ Source listing ↗

Could you build this?

Partial Wrapping vector databases and embedding APIs into a context layer is vibe-codeable, but packaging an enterprise-grade, air-gapped, high-throughput VPC deployment with multiple client SDKs requires specialized systems architecture.

What it would actually take: A real version requires an orchestration platform (e.g., Go/Rust backend) managing embedded vector storage (Qdrant, Milvus, or pgvector), hybrid search (BM25 + dense embeddings), document parsing/chunking pipelines, and automated local LLM/embedding inference (vLLM or llama.cpp) packaged for Kubernetes/Helm in air-gapped VPCs. The hard part is ensuring robust, low-latency semantic search and retrieval accuracy across massive multi-tenant data while supporting on-prem air-gapped setups without internet-dependent cloud services.

Discussion

2 comments analyzed.

Feature requests: Deeper integrations with systems that store unstructured data, Import data from external sources into Skald

Competitors

Other products that read as similar to this one : 275 launches clear the similarity bar, closest 8 shown.

Attention rank: #265 of 276 (itself plus its competitors, highest first, normalized so YC and Product Hunt are compared fairly).

Launched 35 days after the earliest competitor.

The niche, labels, vibe code rating and competitor comparison on this page are analysis by Nicheloom. If you use them, please credit nicheloom.com. How you may use our data.