DEV Community

Jangsik Kim
Jangsik Kim

Posted on Edited on

Filtering RSS & Trend Topics in n8n Before Calling an LLM (Using Jev AI) — Looking for Better Patterns!

Every blog automation starts the same way: a schedule fires, you pull in RSS feeds, Google Trends, news and YouTube titles, and then you hand the whole pile to an expensive LLM and ask "pick three good topics".

That step is where the money goes. Most of the candidates are noise — yesterday's weather, a holiday that already passed, a brand announcement with no search intent behind it. You are paying frontier-model prices to say "no" 80% of the time.

So I moved the "no" upstream.

The pipeline

My n8n workflow runs a Korean-language blog with five channels: product comparisons, IT guides, personal finance, trend analysis, and single-product reviews. It looks like this:

schedule → collect seeds (Google Trends / SNS / News RSS / YouTube)
         → dedupe + normalize
         → [Jev batch prepare] → [Jev call] → [Jev apply]   ← new
         → prompt build → LLM topic generation → validation → sheet
Enter fullscreen mode Exit fullscreen mode

The new part is a three-node detour with a single HTTP Request node inside it. Nothing else in the workflow changed.

What Jev is, in one paragraph

Jev (typesafe/jev-1.14, from TypeSafe AI) is not a chat model. It is a "System One" decision model: you send a state plus a set of questions typed as choice, score, or noul (a yes/no probability), and it returns numbers you can drop straight into an if. Measured in my own testing: 205–539 ms server-side, 0.27–0.92 s round trip. Output tokens are free; you pay per request.

That last property is what makes it usable as a filter: one request can carry up to eight questions and still costs one credit.

Packing eight seeds into one call

The prepare step groups seeds into batches of eight and builds the request body:

const MODEL = 'typesafe/jev-1.14';
const BATCH = 8;        // Jev's per-request question limit
const MAX_SEEDS = 24;   // 24 seeds / 8 = 3 requests = 3 credits per day

const state = batch.map(s => `${s.kind} · ${s.title}`);
const questions = {};
batch.forEach((s, i) => {
  questions['s' + (i + 1)] = {
    type: 'noul',
    instructions: {
      candidate_title: s.title,
      channel_context: CHANNEL_CTX,
      question: 'Is `candidate_title` a topic worth writing today for one of these channels?'
    },
    criteria: {
      true:  'concrete, actually being searched today, has information or purchase intent',
      false: 'vague, trivial, out of season, or no searchable or revenue demand'
    }
  };
});
return [{ json: { body: { model: MODEL, state, questions } } }];
Enter fullscreen mode Exit fullscreen mode

The HTTP Request node is deliberately boring:

POST https://thejevai.com/v1/systemone
Authorization: Bearer {{ $env.JEV_API_KEY }}
Body: {{ JSON.stringify($json.body) }}
Enter fullscreen mode Exit fullscreen mode

Two settings matter more than the URL:

  • Never Error on, onError: continue, three retries. A filter that can stop the pipeline is worse than no filter at all.
  • Seeds are picked round-robin across sources (trend → sns → news → yt), so one noisy feed cannot eat the whole batch.

Applying the answers

The response comes back keyed by question id, with answers.s1.noul holding a probability between 0 and 1. I keep everything at or above 0.45 and re-attach each answer to its original seed.

Two guards, both learned the hard way:

  • A floor on kept seeds. If the filter is too aggressive on a quiet day (kept < 8), the highest-scoring rejected seeds are restored. The filter narrows the funnel; it must not empty it.
  • Fail open. If a batch errors out, those seeds pass through untouched. A dead filter degrades to yesterday's behaviour instead of silently producing nothing.

Real numbers from one morning run: 24 seeds checked, 15 kept, 9 dropped, 3 credits, about 900 ms total, 0 failed batches.

The individual judgments looked about right:

Candidate (translated) noul Verdict
2026 budget wireless keyboard recommendations 0.86 keep
Olive Young global sunscreen ranking (K-beauty) 0.78 keep
Youth savings account guide (Korean gov. scheme) 0.72 keep
iPhone 18 rumors 0.63 keep
Samsung 3nm node status 0.46 borderline keep
Chuseok holiday highway traffic 0.32 drop
Yesterday's Seoul weather 0.11 drop

The two seasonal rejects are the ones I care about most. A plain embedding-similarity filter would happily keep "Chuseok holiday traffic" because it looks like the travel and finance topics that do well. "Is this worth writing today, in this country" is not a similarity question.

Why not embeddings or a small local model?

Honest answer: I have not measured them head to head yet, and that is part of why I am posting this.

My reasoning so far:

  • Embeddings need an anchor and a threshold. Great at dedupe and near-duplicate detection, which I still do separately. Weak at "is this worth writing today", because seasonality and intent are not distances.
  • A local classifier needs labelled data I do not have. At 24 seeds a day, I would be labelling them myself for weeks.
  • A frontier LLM does all of it well and costs roughly two orders of magnitude more per item at this volume.

The decision model sits in between: no labels, no vectors, a plain-language rubric I can edit without retraining, and a sub-second answer. It also decides whether to spend the big-model call, which is where the saving actually comes from.

The cost shape

Because billing is per request and not per token, batching is the whole game:

  • 8 seeds per request x 3 requests a day = 3 credits a day, about 90 a month
  • the free signup credit was 10 (expiring in three days), so this needs a paid tier to become a daily habit ($10 for 100k credits in my case)
  • dropping MAX_SEEDS to 16 halves the daily cost

What I would like to hear from you

If you have built this layer before, I would love your take on:

  1. Threshold tuning. 0.45 came from about 30 samples. Do you drift it per source, or per weekday, or keep one number?
  2. Batch size. Is 8 the sweet spot for you, or does judgment quality degrade before the limit? I have not run smaller batches against the same seeds yet.
  3. Alternatives. Is there a local classifier or an embedding-plus-reranker setup that beats a decision model on 20-30 short texts a day, on a cheap box?
  4. Failure policy. Fail open (mine) or fail closed? I chose open so the blog never misses a day, but that means a broken filter can silently ship bad topics.

I am cross-posting this on Medium and Quora as well, and I will fold the best answers back into the workflow. If you are running something similar, I would rather learn from your scars than collect my own.

Top comments (3)

Collapse
 
agentsearchhq profile image
AgentSearch •

Moving the "no" upstream is the right call. A few cheap filters that usually take out a lot of noise before anything model-shaped sees the seeds:

  • Drop by age first. Most RSS items carry a pubDate, and for a trend channel anything older than a day or two is rarely worth a credit.
  • Keep a sheet of titles you've already written or rejected, normalize them (lowercase, strip punctuation), and skip near-duplicates with a simple token-overlap check. Recurring seeds like annual holidays fall out fast.
  • A small per-channel blocklist (weather, sports scores, "live updates") catches more than you'd expect.
  • Cache the verdict by normalized title for a few days, since the same seed often shows up from Trends and News on back-to-back runs.

Curious whether you've tried asking channel fit and "worth writing today" as two separate questions, or whether one combined yes/no is enough.

Collapse
 
rushdeploy profile image
Rishabh Mishra •

With only about 30 samples, I would keep one threshold and log the score per source instead of drifting it per weekday — at 24 seeds a day, weekday splits are mostly noise. Fail open is the right call for a daily blog, but add a counter that alerts you when a batch fails, so a broken filter does not silently pass everything for a week.

Collapse
 
useaifix profile image
AIFix •

Nice idea moving the "no" upstream. Since the 0.45 cutoff is a fixed number, have you checked how many of the rejected seeds you would have kept after the fact? I'd be curious whether it needs tuning per channel, say finance versus product reviews.