Turn document collections
into knowledge you
can navigate.
SideCloud builds retrieval and knowledge systems for complex documents. Query information, follow relationships, and inspect the sources behind every answer.
What liability limits apply when a supplier exceeds delivery timelines?
"In the event of delay exceeding 10 business days, the supplier shall be liable for liquidated damages not exceeding 0.5% of the total order value per day of delay…"
"The aggregate liability of either party shall not exceed the total contract value for the preceding 12-month period, except in cases of gross negligence…"
"Days 1–10: no penalty; Days 11–20: 0.3%/day; Days 21+: 0.5%/day (aggregate cap per § 12.1 applies)"
Delivery delays beyond 10 business days trigger liquidated damages of up to 0.5% of order value per day (§ 8.41), subject to an aggregate cap equal to the preceding 12-month contract value (§ 12.12). Annex B3 provides a graduated penalty breakdown by delay duration.
Illustration · document query interface
Most tools treat a document as a flat list of text.
Structure is where the meaning lives.
Contracts, regulations, technical standards, and internal knowledge bases are not flat. They have sections that reference other sections, definitions that qualify later clauses, and hierarchies that determine what a passage actually means.
When a retrieval system ignores that structure, it returns text without context. Answers become hard to verify, hard to trust, and hard to act on — especially when the stakes are high.
SideCloud turns document collections into a structured knowledge graph that can be queried, navigated, and verified.
Sections, hierarchies, and cross-references are kept intact when documents are ingested, so retrieval can follow relationships rather than search across isolated fragments.
Every response is connected to the exact passages it drew from. You can verify what the system found, where it came from, and why it was included.
Built for document sets where structure matters — regulatory corpora, legal archives, technical standards, or internal knowledge bases with interdependencies.
Available via API and MCP — compatible with PostgreSQL and any interface that speaks MCP, so it fits into your existing infrastructure rather than requiring a separate deployment.
Where we are
Pre-seed stage, based in Portugal. We have built a working system that turns document collections into a searchable knowledge graph — ingesting structure, extracting relationships, and returning answers with traceable sources.
We are continuing to research and build the next layer of the product, and looking for early partners with real document problems to work with.
Team
Background in AI infrastructure, platform engineering, and SRE. Builds production systems in Python and Rust — the same stack SideCloud runs on.
SideCloud emerged from direct experience with the gap between what RAG products promise and what production retrieval actually requires.
LinkedIn ↗For investors
Raising a pre-seed round to accelerate product development and bring on early design partners. If you are interested in document intelligence infrastructure or European AI, we would like to talk.