Not Everything Needs an LLM
A PostHog newsletter post by Jina Yoon describes how the team uses calibrated decision models, which return typed answers with probabilities, alongside LLMs.
Source-based brief · Editor review not recorded · See checks and limits ↓
One primary source; other links are context, not independent corroboration.
- Primary sourcenewsletter.posthog.com
- Discovery linkHacker News
- Linked in sourcedecision models like Jev
- Linked in sourceReplay Vision
- Linked in sourceIn production
- Linked in sourcewithout being tagged
- Linked in sourcePostHog in Slack
- Linked in sourcemodel router
- Linked in sourceError Tracking
- Linked in sourcehand-labeled data
- Linked in sourceFernando
- Linked in sourcereranked
- Linked in sourceusing a decision model
- Linked in sourcesmarter chart picker
A community post points to the original link. Discussion and upvotes do not establish accuracy.
Open full article view →- Community discussion+5
- Fetched source page+30
- No editor link review+0
- Community link+0
This measures the stored source trail and editorial steps. It does not certify accuracy, peer review, or independent corroboration. Popularity adds no points.
View full scoring →Read the scoring rubric →What the source says
A PostHog newsletter post by Jina Yoon describes how the team uses calibrated decision models, which return typed answers with probabilities, alongside LLMs. Reported uses include gating expensive LLM calls for PII scrubbing, where 98.7% of summaries skip the LLM, routing requests to cheaper models, face-value reranking of search results, and choosing SQL chart types. The post also notes limits: thresholds vary by task and model, and multi-step reasoning and open-ended judgments are better left to LLMs.
Claim explorer
Attribution, source and conditionsSource-stated points from the stored brief. Claim-level assessments have not been recorded.
Source claims not independently checked; figures such as cost multipliers and the 98.7% skip rate are the author's self-reported production numbers from a single company.
This connects a source, an attributed statement and its qualifications. It does not establish that the claim is correct.
Questions worth asking
Open questions for further reportingWhat remains unconfirmed?
Source claims not independently checked; figures such as cost multipliers and the 98.7% skip rate are the author's self-reported production numbers from a single company.
What material was available for this brief?
The recorded brief basis is fetched source page. These points are AINewsIQ paraphrases, not quotations or independent confirmations.
Narrow yes/no decision-model gates let PostHog skip an LLM call for 98.7% of session summaries in its PII pipeline.
Source: newsletter.posthog.com ↗Stored source-based brief · key point 1 · AINewsIQ paraphraseDecision-model thresholds did not carry over between model versions, so the team made them configurable per project and recalibrated.
Source: newsletter.posthog.com ↗Stored source-based brief · key point 2 · AINewsIQ paraphraseChaining decision models into multi-step judgments was operationally costly, flooding the model server and blocking other calls for hours.
Source: newsletter.posthog.com ↗Stored source-based brief · key point 3 · AINewsIQ paraphraseIn an internal test, a decision model on PR findings gave an AUC of 0.52, which the post calls inconclusive.
Source: newsletter.posthog.com ↗Stored source-based brief · key point 4 · AINewsIQ paraphraseLimits from the stored brief
Source claims not independently checked; figures such as cost multipliers and the 98.7% skip rate are the author's self-reported production numbers from a single company.
This is a source-based abstract, not independent validation of the source's claims.
Evidence & sources
These links show where the signal came from and what the source page linked to. A linked page is context, not independent corroboration.
A community post points to the original link. Discussion and upvotes do not establish accuracy.
Timeline & updates
- Signal dated by source or discovery feed
This timestamp places the signal in the radar. It may be the discovery post date rather than the primary page's first publication date.
- Shared brief stored
AINewsIQ prepared one abstract from fetched source page.
Brief history · 1 retained version
These are changes to AINewsIQ’s own publication. Earlier history is unavailable unless recorded here. Complete retained versions are available to the editor.
- Version 1 · published
Source-based brief stored or updated. This does not establish that the publisher changed its page.
Automated record
Source observations · 1 recent captures
A changed capture means the extracted text differed. It can reflect page furniture or extraction changes; it does not establish a substantive correction. Source captures remain private.
Retained capture; no earlier comparison available.
New engagement snapshots do not mean the source itself changed. Source observations and AINewsIQ brief revisions are separate records.
Why this may matter
The post offers practical patterns for cutting LLM cost and latency by filtering and routing requests, with concrete trade-offs around thresholds and error types that teams building AI pipelines can apply.
Limits and open questions
Source claims not independently checked; figures such as cost multipliers and the 98.7% skip rate are the author's self-reported production numbers from a single company.
Organization-specific financial, security, and operational effects require the organization's own context and review.

Join the discussion
Have a different perspective or a question about the evidence? Discussion is free with a verified email account. Every comment is reviewed before publication.
No published comments yet.
New commenters: 2 submissions per 24 hours, at least 5 minutes apart, until 3 are approved. Discussion guidelines and moderation →