1
Connect your sources
Add RSS feeds, newswires, your CMS, social inboxes and briefing documents. Ottili LD3 ingests them with provenance and keeps the original attribution.
Ottili LD3
Content automation for news, blogs, shops and social
Ottili LD3 ingests and verifies your sources, extracts the facts, drafts independently, routes through reviews and approvals, publishes to the right channels, and learns from what performs. One pipeline, from source to published package.
Beta
How it works
From verified source to published package
Ottili LD3 is built for the full content loop: intake, verification, fact extraction, generation, media planning, reviews, packaging, approvals, publishing and performance. Most steps run automatically — only source and channel setup, plus reviews and publish approvals, need you.
2
Verify and extract facts
Ottili LD3 verifies each source, surfaces conflicts, and turns verified material into a structured fact base of claims, entities, dates and figures.
3
Generate and review
Drafts are written from the fact base, reviewed by humans and AI, and assembled into reusable content packages with shared metadata.
4
Approve, publish and learn
Approval gates hold external posts until confirmed, packages publish to the right channels, and analytics feed the next planning cycle.
The loop
A source becomes a durable, publishable package
Ottili LD3 keeps a clear line between ingestion and authoring: verified facts feed independent drafts, and every external post waits for an approval before it ships.
Ottili LD3 turns verified sources into published content across your news site, blog, shop and social channels. Some steps run automatically; others need a connected source or channel, or a named approval before they proceed. Here is exactly how the pipeline works.
How each step runs
- AutomaticLD3 runs this step for you once the pipeline is configured.
- ConfigurationYou connect a source, channel or brand kit before this step runs.
- ApprovalA named person reviews and confirms before this step proceeds.
- Step 1Source intakeConfiguration
- Step 2VerificationAutomatic
- Step 3Fact extractionAutomatic
- Step 4GenerationAutomatic
- Step 5Media planningAutomatic
- Step 6ReviewsApproval
- Step 7Content packagesAutomatic
- Step 8ApprovalsApproval
- Step 9PublishingConfiguration
- Step 10PerformanceAutomatic
- Step 11LearningAutomatic
Example content package
Quarterly logistics cost outlook
- News800-word explainerPublished
- BlogSEO long-form + FAQPublished
- ShopCategory page copyApproved
- Social4-post thread + carouselIn review
- NewsletterShort recap blurbQueued
One package,news · blog · shop · social · newsletter·shared facts, one approval trail
Step 1
Source intake
You see
You connect RSS feeds, newswires, your CMS, social inboxes and briefing documents — or set a topic to monitor. Each source keeps its original link and attribution.
Behind the scenes
LD3 ingests sources with provenance metadata so every later draft can trace a claim back to where it came from — no anonymous scraping. Topic inputs seed the same monitored source set.
This step runs as: Configuration. You connect a source, channel or brand kit before this step runs.
Step 2
Source verification
You see
Before anything is used, sources are checked for credibility, freshness and internal conflict. Low-trust or conflicting sources are flagged, not silently used.
Behind the scenes
A verification pass scores each source and surfaces conflicts between materials, so generation only builds on sources you can stand behind.
This step runs as: Automatic. LD3 runs this step for you once the pipeline is configured.
Step 3
Fact extraction
You see
Verified material is reduced to structured claims — entities, dates, figures and quotes — that drafts can cite instead of paraphrase from memory.
Behind the scenes
LD3 extracts a fact base from verified sources so generated copy references real, attributable facts rather than plausible-sounding fabrications.
This step runs as: Automatic. LD3 runs this step for you once the pipeline is configured.
Step 4
Independent generation
You see
Drafts are written from the verified fact base, not the raw source text, keeping a clear line between ingestion and authoring.
Behind the scenes
Generation works off the structured fact base so the output is independently composed — it explains and reorganises, rather than copying source phrasing.
This step runs as: Automatic. LD3 runs this step for you once the pipeline is configured.
Step 5
Media planning & assets
You see
From one draft, LD3 produces the right media for each channel — social cutdowns, image briefs, short video scripts and SEO metadata — without re-authoring from scratch.
Behind the scenes
Media planning maps the approved draft to each channel's format and brand kit, so a single story becomes a consistent set of on-brand assets automatically.
This step runs as: Automatic. LD3 runs this step for you once the pipeline is configured.
Step 6
Human & AI reviews
You see
Reviewers see each draft next to its source facts and can comment, reject or request changes before anything ships.
Behind the scenes
Reviews pair human editors with AI review checks (tone, accuracy, source alignment) so nothing leaves without a recorded review state.
This step runs as: Approval. A named person reviews and confirms before this step proceeds.
Step 7
Content packages
You see
Approved pieces are assembled into reusable packages — article, social cutdowns, newsletter blurb and shop copy — sharing one set of metadata.
Behind the scenes
LD3 groups related assets into a single content package so a story stays consistent across every channel it touches.
This step runs as: Automatic. LD3 runs this step for you once the pipeline is configured.
Step 8
Approval gates
You see
Publication holds external posts until a named approver confirms. The decision is recorded on the package.
Behind the scenes
Approval gates keep external publishing auditable: who approved, when, and on which channel — matching the Ottili ONE approval system.
This step runs as: Approval. A named person reviews and confirms before this step proceeds.
Step 9
Multi-channel publishing
You see
Packages push to the right destination — your news site, blog, shop product pages or social accounts — through Ottili ONE-connected channels.
Behind the scenes
Publishing routes each asset to its mapped channel and tracks delivery status, so a package is one action, not ten manual posts. Connected channels are a prerequisite.
This step runs as: Configuration. You connect a source, channel or brand kit before this step runs.
Step 10
Performance analytics
You see
Reach, engagement and conversion are measured per package and per channel, giving you a clear read on what worked.
Behind the scenes
LD3 aggregates performance back to the content package and source, so the next planning cycle starts from evidence rather than intuition.
This step runs as: Automatic. LD3 runs this step for you once the pipeline is configured.
Step 11
Learning loop
You see
What performed and what was rejected teaches topic selection, tone and source weighting over time.
Behind the scenes
The learning loop feeds planning: stronger sources rise, weak angles fade, and future packages inherit what already proved effective.
This step runs as: Automatic. LD3 runs this step for you once the pipeline is configured.
Capabilities
Built for trustworthy content, not demo theater
Ottili LD3 joins source intake, verification, fact extraction, independent generation, reviews, packaging, approvals, publishing and analytics into one product surface — with the Ottili ONE approval system underneath.
Source intake with provenance
Connect RSS feeds, newswires, your CMS, social inboxes and briefing docs. Every source keeps its original link and attribution so drafts can trace claims back to where they came from.
Source verification
Each source is checked for credibility, freshness and internal conflict before it feeds generation. Low-trust or conflicting sources are flagged, never silently used.
Fact extraction
Verified material is reduced to structured claims — entities, dates, figures and quotes — so drafts cite real, attributable facts instead of paraphrasing from memory.
Independent generation
Drafts are written from the verified fact base, not the raw source text, keeping a clear separation between ingestion and authoring. Output explains and reorganises rather than copying source phrasing.
Media planning & assets
From one draft, Ottili LD3 produces the right media for each channel — social cutdowns, image briefs, short video scripts and SEO metadata — without re-authoring from scratch.
Human & AI reviews
Reviewers see each draft beside its source facts and can comment, reject or request changes. AI review checks cover tone, accuracy and source alignment before anything ships.
Content packages
Approved pieces are assembled into reusable packages — article, social cutdowns, newsletter blurb and shop copy — that share one set of metadata and stay consistent across channels.
Approval gates
Publication holds external posts until a named approver confirms. The decision is recorded on the package, matching the Ottili ONE approval system used across products.
Multi-channel publishing
Packages push to the right destination — your news site, blog, shop product pages or social accounts — through Ottili ONE-connected channels, with delivery status tracked per asset.
Analytics
Reach, engagement and conversion are measured per package and per channel, giving a clear read on what worked and feeding the next planning cycle with evidence.
Learning loop
What performed and what was rejected teaches topic selection, tone and source weighting over time, so future packages inherit what already proved effective.
Channels
One package, every surface
The same verified fact base powers news, blogs, shops and social — each channel gets the right cut of the package, all sharing one approval trail.
News
Turn verified sources into accurate explainers and breaking-news briefs with citations and a clear review trail.
- 1Ingest wires, press releases and official sources
- 2Extract dated facts and quotes
- 3Draft with source links attached
- 4Approve and publish to the news site
Blogs
Produce SEO long-form articles and FAQ content from the same fact base, consistent with your editorial voice.
- 1Plan topics from performance and gaps
- 2Generate long-form with internal links
- 3Human review for tone and accuracy
- 4Schedule and publish to the blog
Shops
Write and update product, category and campaign copy that stays aligned with verified source material.
- 1Pull verified specs and claims
- 2Generate product and category copy
- 3Route through approval gates
- 4Push to shop product pages
Social
Cut a single package into on-brand social posts and threads for each network, with approval before they go live.
- 1Derive post cutdowns from the package
- 2Adapt tone per network
- 3Hold for named approver
- 4Publish to connected accounts
Quality & originality
Built to keep content accurate and original
Ottili LD3 is designed to limit the usual AI pitfalls — hallucination, unsupported claims, source copying and low-quality output — through fact extraction, a per-claim source matrix, independent generation, a separate review, a rework loop and an approval gate. It helps, but it does not guarantee error-free publishing.
AI-written content can still hallucinate, make unsupported claims, copy source phrasing or read as low quality. Ottili LD3 is built around a set of controls that work together to limit each of those risks — from the first verified fact to the final approval. Here is how the quality and originality pipeline holds up.
Fact extraction into a structured base
Verified sources are reduced to structured claims — entities, dates, figures and quotes — so drafts cite real, attributable facts instead of plausible-sounding fabrications.
This control mainly helps reduce:
Helps reduce
HallucinationUnsupported claims
A source matrix for every claim
Each claim in the fact base keeps its supporting source link and attribution. Writers and reviewers can see which source backs each statement, so nothing ships without a reference it can stand behind.
This control mainly helps reduce:
Helps reduce
Unsupported claimsHallucination
Independent generation from the fact base
Drafts are composed from the verified fact base, not copied from raw source text. The output explains and reorganises rather than mirroring source phrasing, which helps limit verbatim source copying.
This control mainly helps reduce:
Helps reduce
Source copyingLow-quality output
A separate human and AI review
Reviewers assess each draft beside its source facts and can comment, reject or request changes. AI review checks cover tone, accuracy and source alignment before anything moves to approval.
This control mainly helps reduce:
Helps reduce
Low-quality outputUnsupported claims
A rework loop on rejection
When review finds a wrong figure, a gap or weak writing, the draft returns to generation with the feedback attached. It is rewritten until it clears review — it is never quietly published as-is.
This control mainly helps reduce:
Helps reduce
Low-quality outputUnsupported claims
An approval gate before publishing
External posts wait for a named approver. The decision is recorded on the package, so public content only ships after a person confirms it is accurate and on-brand.
This control mainly helps reduce:
Helps reduce
Unsupported claimsLow-quality output
Example source matrix
Every claim points back to a source
| Claim | Sources | Status |
|---|---|---|
| Quarterly logistics costs rose 4.2% | Industry wireFinance releaseTwo independent sources agree; figure quoted, not paraphrased. | Supported |
| New regional warehouse opens in Q1 | Press releaseSingle primary source; flagged for a second confirmation before publish. | Supported |
| Free shipping for all customers this quarter | No sourceNo source backs this claim, so it is held for review instead of drafted as fact. | Held for review |
Example rework diff
Rejected drafts are rewritten, not shipped
Rejected in review
Logistics costs are way up across the board this year.
After rework
Logistics costs rose 4.2% in Q3, led by fuel and regional handling fees.
Review rejected the vague claim; generation rewrote it from the verified figure with a cited source.
No automated system removes every risk. LD3 is designed to reduce hallucination, unsupported claims, source copying and low-quality output — but it does not guarantee that published content is error-free. Final accuracy still depends on the sources you connect and the people who review and approve each package.
Focused use case
A newsroom built on the same pipeline
Local publishers and content teams can run the whole loop as a focused newsroom: source monitoring, topic queues, fact quality, publishing levels and editorial review. The pilot starts from one verified fact base.
Pricing
Ottili LD3 pricing, sourced from central Ottili ONE plans
Ottili LD3 follows the same plans as Ottili ONE. The matrix packages Ottili LD3 across sources, articles & content packages, channels, websites, review depth, publishing level, team workflows, add-ons and AI credits — with no unlimited expensive inference. Newsrooms and agencies read the same packaging.
| Ausgewählter Tarif | Free€0 | Starter€39/mo | Pro€299/mo | Business€499/mo |
|---|---|---|---|---|
Verbundene QuellenLiveVerbinde RSS-Feeds, Nachrichtenagenturen, dein CMS, Social-Postfächer und Briefing-Dokumente. Jede Quelle behält Original-Link und Quellenangabe, sodass Entwürfe Aussagen zurückverfolgen. | Nicht enthalten | Nicht enthalten | Nicht enthalten | Enthalten |
Artikel & Content-PaketeLiveFreigegebene Beiträge werden zu wiederverwendbaren Paketen — Artikel, Social-Kürzungen, Newsletter-Text und Shop-Copy — mit gemeinsamen Metadaten. Die Erzeugung wird über Ottili AI-Credits abgerechnet. | Nicht enthalten | Nicht enthalten | Nicht enthalten | Credits |
VeröffentlichungskanäleLiveLiefere Pakete über Ottili ONE-verbundene Kanäle an News, Blog, Shop und Social — mit Lieferstatus pro Asset. | Nicht enthalten | Nicht enthalten | Nicht enthalten | Enthalten |
Verbundene WebsitesLiveBetreibe eine verifizierte Faktengrundlage über die Websites und News-Seiten, die du veröffentlichst, und halte die Copy pro Kanal und Sprache konsistent. | Nicht enthalten | Nicht enthalten | Nicht enthalten | Enthalten |
Review-TiefeLiveMenschliche und KI-Review liegen neben jedem Entwurf samt Quellenfakten. Reviewer kommentieren, lehnen ab oder fordern Änderungen, bevor etwas veröffentlicht wird. | Nicht enthalten | Nicht enthalten | Nicht enthalten | Enthalten |
VeröffentlichungsstufeLiveFreigabe-Gates halten externe Beiträge, bis ein benannter Freigeber bestätigt; die Entscheidung wird am Paket festgehalten — passend zum Ottili ONE-Freigabesystem. | Nicht enthalten | Nicht enthalten | Nicht enthalten | Enthalten |
Team-WorkflowsLiveRollen, Berechtigungen, gemeinsame Workspaces und ein Audit-Trail folgen der Teammitgliedschaft des Tarifs über den gesamten Redaktionsloop. | Nicht enthalten | Nicht enthalten | Nicht enthalten | Enthalten |
Add-onsGeplantOptionale Add-ons — zusätzliche Media-Planung, Übersetzungs-Packs, SEO-Erweiterungen — sind geplant und werden transparent auf den Tarif aufgeschlagen, nie versteckt. | Geplant | Geplant | Geplant | Geplant |
AI-CreditsLiveJede KI-Aktion — Erzeugung, Faktenextraktion, Media-Planung, KI-Review — wird über Ottili AI-Credits abgerechnet. Du behältst die Kostenkontrolle über dein Credit-Budget. | Nicht enthalten | Nicht enthalten | Nicht enthalten | Enthalten25,000 Credits / Monat |
Ausgewählter Tarif: Pro · €299/mo
Was jeder LD3-Tarif enthält — Enthalten · Credit-basiert (begrenzt durch Ottili AI-Credits) · Geplant · Nicht enthalten
- Verbundene QuellenLive
- Verbinde RSS-Feeds, Nachrichtenagenturen, dein CMS, Social-Postfächer und Briefing-Dokumente. Jede Quelle behält Original-Link und Quellenangabe, sodass Entwürfe Aussagen zurückverfolgen.
- Nicht enthalten
- Artikel & Content-PaketeLive
- Freigegebene Beiträge werden zu wiederverwendbaren Paketen — Artikel, Social-Kürzungen, Newsletter-Text und Shop-Copy — mit gemeinsamen Metadaten. Die Erzeugung wird über Ottili AI-Credits abgerechnet.
- Nicht enthalten
- VeröffentlichungskanäleLive
- Liefere Pakete über Ottili ONE-verbundene Kanäle an News, Blog, Shop und Social — mit Lieferstatus pro Asset.
- Nicht enthalten
- Verbundene WebsitesLive
- Betreibe eine verifizierte Faktengrundlage über die Websites und News-Seiten, die du veröffentlichst, und halte die Copy pro Kanal und Sprache konsistent.
- Nicht enthalten
- Review-TiefeLive
- Menschliche und KI-Review liegen neben jedem Entwurf samt Quellenfakten. Reviewer kommentieren, lehnen ab oder fordern Änderungen, bevor etwas veröffentlicht wird.
- Nicht enthalten
- VeröffentlichungsstufeLive
- Freigabe-Gates halten externe Beiträge, bis ein benannter Freigeber bestätigt; die Entscheidung wird am Paket festgehalten — passend zum Ottili ONE-Freigabesystem.
- Nicht enthalten
- Team-WorkflowsLive
- Rollen, Berechtigungen, gemeinsame Workspaces und ein Audit-Trail folgen der Teammitgliedschaft des Tarifs über den gesamten Redaktionsloop.
- Nicht enthalten
- Add-onsGeplant
- Optionale Add-ons — zusätzliche Media-Planung, Übersetzungs-Packs, SEO-Erweiterungen — sind geplant und werden transparent auf den Tarif aufgeschlagen, nie versteckt.
- Geplant
- AI-CreditsLive
- Jede KI-Aktion — Erzeugung, Faktenextraktion, Media-Planung, KI-Review — wird über Ottili AI-Credits abgerechnet. Du behältst die Kostenkontrolle über dein Credit-Budget.
- Nicht enthalten
Enthalten · Credit-basiert (begrenzt durch Ottili AI-Credits) · Geplant · Nicht enthalten
Die Preise stammen aus der zentralen Ottili ONE-Preisquelle. KI-Nutzung ist credit-basiert und begrenzt; es gibt keine unbegrenzte teure Inferenz. Eine Belastung erfolgt niemals ohne ausdrückliche Bestätigung.
Ready to publish
Use Ottili LD3 as its own product or inside Ottili ONE
Open Ottili LD3 as a standalone product, or connect it to your Ottili ONE workspace. Either way it owns the pipeline: intake, verification, fact extraction, generation, reviews, packaging, approvals, publishing and analytics.