Patent Agent

One engine.
Five jurisdictions. Ten AI endpoints.

Patent Agent is where Patent & R&D Operations' AI actually runs — jurisdiction-aware filing forms for TR/EPO/US/WIPO, and ten dedicated reasoning endpoints for collision risk, prior art, claim charts, valuation, and more.

5 Jurisdictions 10 AI Endpoints Two-Stage Pipeline Model Routing
5 Jurisdictions

Filing forms that know
whose rules apply

Claim structure and required abstract fields differ by patent office. Select a country and the form itself changes — field names, filing fees in the right currency, and the statute that governs it.

  • Turkey (TÜRKPATENT) — 6769 Sınai Mülkiyet Kanunu, filing fee in ₺
  • EPO member states (16 countries) — European Patent Convention Rules, filing fee in €
  • United States (USPTO) — 35 U.S.C., provisional + first-inventor-to-file dates, filing fee in $
  • WIPO / PCT — international filing, designated states, filing fee in CHF
Open Patent Registry Setup
totalapp.app / patent-registry-setup
Jurisdiction: EPO (European Patent Convention)
IPC ClassG06F 17/30
Priority Date2026-01-14
Filing Fee€1,240
Fields regenerate per jurisdiction
10 AI Endpoints

One reasoning engine
per screen, on purpose

Every AI-calling screen in Patent & R&D Operations has its own dedicated endpoint — collision risk, prior art novelty, claim charts, FTO clearance, literature synthesis, grant compliance, license audits, and IP valuation. Each one's prompt and output schema stay independent, so a change to one screen's reasoning can never silently affect another's.

  • Collision & Prior Art — risk level, overlap summary, novelty classification
  • Claims & FTO — element-by-element claim charts, market clearance status
  • Literature & Grants — consensus level, feasibility scoring, compliance audits
  • Licensing & Valuation — clause audits, fair market value, royalty rates
Explore Patent & R&D Operations
/api/patent-engine/*
collision-overlap-analyzecritical_overlap
prior-art-searchblocking_art
fto-risk-synthesizecleared
ip-valuation-analyze$2.4M
10 endpoints · same routing, own schema
Two-Stage Pipeline

Find similar first.
Reason deeply second.

Nearly every screen runs the same two-stage flow. Stage 1 narrows a target spec, claim, or query down to its closest matches using local embeddings — that's Knowledge Engine, always on your own machine. Stage 2 is Patent Agent reasoning over that shortlist.

  • Stage 1Knowledge Engine ranks candidates by meaning, entirely on your local Ollama instance
  • Stage 2 — Patent Agent reasons over the top matches, routed by your Settings → Agentic choice
  • Never touches the server when Stage 1 or Stage 2 runs on Ollama/local-LLM/web-LLM
  • Every server call logged — duration, model, screen name — visible in Settings → Analytics
See Knowledge Engine
patent-collision-scanner
1 Knowledge Engine — top 3 of 5 competitor patents
2 Patent Agent — collision-overlap-analyze
Same pattern across all 9 Patent & R&D screens
Model Routing

Cohere, Anthropic,
or your own CLI

Every one of Patent Agent's ten endpoints applies the same routing logic once a request reaches the server — which only happens when you've chosen API or Local CLI in Settings → Agentic. Ollama, local-LLM, and web-LLM never send a request here at all.

  • Cohere — when apiMode is set to Cohere, routes to command-r7b-12-2024
  • Anthropic API — default when an API key is configured, uses claude-sonnet-4-6
  • Local Claude CLI — automatic fallback with no API key configured
  • Every call logged to Event & Analytics Engine with duration, model, and screen name
Read the Full API Docs
Settings → Agentic
// apiMode routing
'cohere' → command-r7b-12-2024
key set → claude-sonnet-4-6
no key → local claude --print
Ollama/local-LLM/web-LLM never reach this server