How We Cut Newsletter Production Time by 67% with SCOVA AI
This case study details how we cut production time for the Briefinn newsletter by 67%. By shifting from scattered research to an AI-driven workflow, we transformed our editorial…
Roberto de Rosa · 2026-09-11 · 7 min
Anyone who produces newsletters, press reviews, or informational briefs knows the paradox: credibility requires many sources, but more sources mean more hours spent on scouting, reading, and synthesizing. The result is that curation quality becomes inversely proportional to the sustainability of the process.
In the era of generative AI, the promise was to solve this bottleneck, but most tools stopped at summarization. They're great for summarizing what you've already found, but useless for finding it in the first place. The real leap happens when the prompt becomes the interface to discover, filter, and package news in a single flow—a paradigm shift we explored in our article on conversational interfaces for news feeds.
The Curated Newsletter Problem: Disinformation and Overload
The modern curator lives in a state of constant tension. Relying on a few established sources creates bias and geographic blind spots; opening up monitoring to hundreds of publications leads to paralysis and unsustainable production times. Both extremes damage the newsletter: the first becomes predictable, the second never gets published on time.
To this, we add a hidden cost: generic summarization tools like ChatGPT or Claude are powerful on the text you provide, but they don't solve the most time-consuming part of the job—the discovery of relevant sources that aren't already in your bookmarks. Summarizing is easy; selection is where editorial value is created.
'The useful concept here is that of informational "right-sizing": neither the echo chamber of three recurring publications nor the background noise of an undifferentiated aggregator. A valuable newsletter lives in that intermediate zone where coverage is broad, but the filter is strict.
Anatomy of Briefinn's Manual Workflow
Producing Briefinn—Minnovo's informational newsletter—involved four distinct phases.
Definition of the theme and editorial angle: what to include, what to exclude, and what perspective to take.
Source scouting: the most costly phase, as it required going far beyond the usual publications.
Selective reading: discarding nine out of every ten articles.
Editorial synthesis: transforming the selected material into the final brief.
We measured this workflow precisely: approximately 3 hours for each issue, with more than half of that time spent on the first two phases. The limitation wasn't reading speed but the human capacity to monitor sources in multiple languages and geographies simultaneously—a structural problem that recurs in all professional media monitoring and fact-checking workflows.
No professional, however expert, can manually monitor information at a multilingual, multi-geographic scale. It's a structural problem, not a skills gap.
The Result: -67% in Scouting Time with a Hybrid Approach
When we integrated Scova AI into the Briefinn workflow, the timelines changed radically:
| Phase | Before | After |
| :--- | :---: | :---: |
| Scouting and Synthesis | ~3 hours | ~1 hour |
A 67% reduction—not a marginal efficiency gain, but a shift in the operating model.
'The important point is that we deliberately chose a hybrid approach: we still manually select the news to include in Briefinn from the Scova feed. This is because the tool is currently in Beta, and we want to maintain full editorial control during this phase.
When Scova exits Beta, it will be able to manage the entire flow autonomously—with processing times of less than 1 minute to produce the complete brief. The current hybrid approach is therefore a conscious transitional choice, not a permanent limitation.
From Feed Reader to Prompt: The Generational Leap
Classic RSS readers like Feedly or Inoreader introduced a first level of automation, but their logic remains configurational: the user must know in advance which feeds to follow, build folders, and set up boolean rules. This works if you already have a complete mental map of your information domain; it fails when you want to explore a new one.
A natural language interface reverses this dynamic. Instead of configuring feeds, you describe interests. Instead of maintaining lists, you delegate the discovery of sources you didn't know existed to the system. It's the difference between using a boolean search engine and asking a context-aware assistant a question—a logic that is particularly effective when monitoring vertical niches poorly covered by mainstream feeds.
The second enabling element is conversational iteration: if the system includes off-target articles, you tell it in plain language, and the filter updates without you having to rebuild complex queries from scratch. This drastically reduces the friction of the initial setup, which is where most users abandon traditional tools.
Manual vs. AI-driven Workflow Comparison (Briefinn Case)
| Characteristic | Manual Workflow | AI-driven Workflow (Hybrid) |
| :--- | :--- | :--- |
| Source Scouting | Time-consuming, limited by human knowledge | Automatic across 150,000+ global sources |
| Content Filtering | Subjective, based on keywords and bookmarks | Prompt-based, semantic, iterative |
| Scouting Time | ~3 hours per issue | ~1 hour (-67%) |
| Coverage | Limited to known sources and languages | Multilingual, multi-geographic |
| Final Selection | Manual | Manual (Hybrid) → Automatic (Post-Beta) |
| Future Time (Post-Beta) | — | < 1 minute for the entire flow |
How Briefinn's Current Workflow Operates
The process we've structured with Scova AI consists of three steps that mirror the human editorial process but significantly compress the time required.
flowchart TD
A[Define Theme & Angle] --> B{AI Filters: Prompt-Based News Discovery}
B --> C[Multi-Source Search — 150k+ sources, 89 languages]
C --> D{Semantic Clustering & Article Ranking}
D --> E[Manual Feed Selection — Hybrid Approach]
E --> F[Produce & Send Briefinn]
Phase 1 — Prompt Definition
We describe the issue's theme, exclusions, geographic areas, and editorial angle in natural language. For Briefinn, a typical prompt is:
"News about innovation from around the world, excluding product reviews and announcements."
Scova AI automatically identifies the best sources and keywords to track, without complex configurations.
Phase 2 — Search and Clustering
The system draws from over 150,000 sources across 206 countries and 89 languages. Semantic clustering identifies related news, deduplicates it, and ranks it by relevance to the prompt. This entire process takes less than 60 seconds.
Phase 3 — Selection (Hybrid, for now)
At this point, the Minnovo team scrolls through the feed and manually selects the articles to include in Briefinn. This is a deliberate choice we've made while Scova is in Beta: we want to maintain fine-grained editorial control over each issue. Once the tool exits Beta, this phase can be automated—reducing the entire production cycle to under a minute.
Reliability and Editorial Control
Total automation is an editorial anti-pattern. Briefinn is a newsletter signed by Minnovo, and no issue can contain false or poorly contextualized news. That's why fact-checking is part of the workflow, not an afterthought.
In Scova AI, every news item can be verified with a click and receive a reliability score on four levels: true, partially true, false, or unverifiable. When a story warrants a deeper look, the Deep Research feature—built by integrating Perplexity into our semantic clustering—aggregates related sources and produces a vertical synthesis in seconds. It's the same logic we apply to AI-assisted investigative journalism.
The curator always has the final word: they can add or remove sources, exclude individual articles, and modify the prompt. The AI is an assistant that amplifies editorial judgment, not a replacement for it.
Curation as a Competitive Advantage
When the production time for a newsletter drops by 67%, the interesting question isn't how much you save, but what you do with the reclaimed hours. You could increase the publishing frequency, expand the number of vertical briefs, or invest the freed-up time in deeper analysis and original content.
In this framework, Briefinn ceases to be a fixed cost to be minimized and becomes a strategic lever for Minnovo's positioning. The speed of curation is directly proportional to perceived relevance—and with a scalable process, editorial quality no longer needs to be sacrificed at the altar of operational sustainability.
The competitive advantage isn't in the AI itself, but in the discipline with which it is integrated into a process that remains, at its core, profoundly human.