Comparative study

Alhena vs. Sierra: policy-resolution comparison

Name: Alhena vs. Sierra: policy-resolution comparison
Creator: Alhena Research Lab
Published: 2026-09-23T00:59:21.323Z

Comparison assembled from published evaluations; no new conversations were run for this comparison.

Published Sep 23, 2026. Captured Sep 21–23, 2026. Commissioned by Alhena Research Lab using the policy-resolution-v1 method.

10storefront deployments

100conversations captured

100conversations judged

10guardrail tests

917checkpoints judged and audited

Key findings

Shopping composite, out of 100: Sierra 59, Alhena 68.4.
Support composite, out of 100: Sierra 61.6, Alhena 81.4.
Overall composite, the mean of both lanes: Sierra 60.3, Alhena 74.9.
Average time to a full answer: shopping Sierra 8.5 s, Alhena 11.8 s; support Sierra 8.4 s, Alhena 8.7 s.
Scores cover the AI agent in each store's on-site chat widget across the 10 storefront deployments captured Sep 21–23, 2026. They do not cover other products either vendor sells, every deployment, or an overall vendor ranking. Alhena operates Alhena Research Lab.

Results

Scores out of 100 for the storefronts in this study. Each bar splits a lane composite into the points resolution, quality and speed add.

Sierra60.3overall composite

Alhena74.9overall composite

ShoppingComposite / 100

Sierra8.5 s to a full answer
59
Alhena11.8 s to a full answer
68.4

SupportComposite / 100

Sierra8.4 s to a full answer
61.6
Alhena8.7 s to a full answer
81.4

Resolution
Quality
Speed

Score breakdown

Each measure out of 100 unless marked, by lane and provider.
Measure	Shopping		Support
Measure	Sierra	Alhena	Sierra	Alhena
Composite	59	68.4	61.6	81.4
Resolution	61.1	64.8	40	72.8
Quality	48	83	86	95
Speed score	71.1	53.7	71.6	70
Average full answer	8.5 s	11.8 s	8.4 s	8.7 s
Storefronts included	5 of 5	5 of 5	5 of 5	5 of 5
Conversations included	25 of 25	25 of 25	25 of 25	25 of 25

Full evidence coverage

Shopping evidence coverage
Company	Checkpoints: planned / attempted / observed	Recorded submitted / assessed / unassessable	Attained / unverified	Contexts: included / excluded	Stores: included / excluded	Quality eligible: contexts / stores	Captures: original / repaired
Sierra	250 / 183 / 183	183 / 183 / 0	104 / 38	25 / 0	5 / 0	20 / 4	24 / 1
Alhena	250 / 242 / 242	242 / 242 / 0	158 / 64	25 / 0	5 / 0	25 / 5	23 / 2

Support evidence coverage
Company	Checkpoints: planned / attempted / observed	Recorded submitted / assessed / unassessable	Attained / unverified	Contexts: included / excluded	Stores: included / excluded	Quality eligible: contexts / stores	Captures: original / repaired
Sierra	250 / 242 / 242	242 / 242 / 0	96 / 137	25 / 0	5 / 0	24 / 5	25 / 0
Alhena	250 / 250 / 250	250 / 250 / 0	182 / 63	25 / 0	5 / 0	25 / 5	19 / 6

Policy-compliant resolution · public-session scope: Verified correct answer or policy-prescribed next step, averaged equally by conversation and storefront. This does not measure downstream case completion.
Quality: The original fixed answer-quality rubric. Unchanged captures retain their original judgments; repaired captures receive a fresh judgment and audit.
Full-answer speed: Average time to the complete answer, scored 100 at 3 seconds, zero at 22 seconds, bounded between 0 and 100.
Composite: Shopping: 40% policy-compliant resolution, 35% quality, 25% speed. Support: 50% policy-compliant resolution, 40% quality, 10% speed.
Overall composite: Equal mean of the two unrounded lane composites; only available when both lanes meet coverage floors.

This study covers the chat widget only

Studies test one form factor: the AI agent in each store’s on-site chat widget. Other products the vendors sell are outside its scope.

Read this before the scores

Policy-compliant resolution can include a verified policy-required next step. It does not establish actual refunds, completed account actions or human-resolved orders. This is a separately versioned Alhena Research Lab method, not an independent certification or the Gorgias leaderboard.

Alhena commissioned and operates this study. Selected storefront deployments do not establish a universal provider ranking or independent certification.
This evaluation used the previously published method, frozen before capture and judging.
The sessions were logged out. No real order, refund, account change or completed human resolution was independently verified.
Unknown delivery or actor attribution remains unassessable. Unsent questions are not successes or failures. Conditional scores must be read with planned and assessed coverage.
Unresolved capture failures pause publication; completed studies do not represent every deployment of a provider.
1 earlier submission attempt(s) preceded operator-corrected captures. They are retained in the repair provenance and excluded from the selected-capture scoring denominators.
This methodology was developed after reviewing the original automation results, then frozen before the new policy-resolution judgments.
All ten JSHealth core contexts were excluded because a generative actor and delivery could not be confirmed. Two additional support contexts stopped at login gates.
Two retained dryrobe shopping contexts have only six measured replies. Twelve cause-selected capture repairs contain all 120 submitted questions; every repaired outcome is retained.
Merchant policy conflicts and gaps are retained. Product efficacy, review authenticity and downstream cart or refund completion were not independently established.
The repaired captures have gaps in live resource telemetry. Terminal state and artifact hashes were verified; continuous resource observations are not claimed.
Private cart, session or customer links are redacted from displayed evidence and quotations. Judgments used the unchanged original captures; these display redactions do not change scores.
Derived from separate published runs under policy-resolution-v1. Original capture dates and coverage are retained; execution environments may differ.

How the scores were checked

This comparison retains the original judgments and audit coverage of each source study. No new judgment or audit is claimed for this derived comparison.

Read the versioned methodology

What changed from the source study

Replaces the source automation classifier with policy-compliant resolution.
Every eligible checkpoint receives two fresh blind judgments; attainment requires agreement and valid evidence.
New automated captures use the frozen protocol; historical study scores remain unchanged.

Audit notes (9)

Provider names were masked where practicable; full anonymity is not claimed.
Each new quality judgment has a separate adversarial audit. The quality audit sees the primary verdict; the policy-resolution audit does not.
Retained original quality judgments have separate historical audit coverage. The repair-quality audit sees the primary quality verdicts and full captured replies; it is distinct from the blind policy-resolution audit.
One initial scoring attempt failed in the response decoder with zero accepted decisions; upstream completion is unknown. The output was unavailable, and the documented infrastructure repair did not select among verdicts.
The first blind policy-resolution audit exhausted its 8,000-token per-model-turn output allowance without a structured verdict. Its unusable output was preserved and the identical input was audited once more under the reviewed 16,384-token per-model-turn allowance.
A later blind audit failed after about 602 seconds without a returned verdict. This is consistent with the original 600-second local timeout, but the inner error was not retained and provider completion is unknown. The unusable attempt was preserved and the identical input was retried once under a uniform 1,200-second future execution limit with a 1,250-second outer transport limit. All 36 previously successful stage outputs remain unchanged under their original timeout contracts.
The exact successful initial primary judgment for 30 checkpoints is retained with an 8,000-token per-model-turn allowance. Every subsequent policy-resolution and repair-quality call uses 16,384 per model turn, with the same frozen prompts, model, effort and criteria. New calls use fresh independent sessions under reviewed pool releases with 2 and 4 active slots; 4 distinct isolated slots were used. A CLI session can contain multiple internal steps, so its total output can exceed the per-turn allowance.
The original quality audit sampled only Alhena records. New policy-resolution audits cover both providers, but this does not retroactively expand historical quality-audit coverage.
Source methods: sierra-ai-98c6f6e402-9cba991e (712b1e5753cd9b871881d328d79ecb20ccbb279c045fe9233891c6805441133b); alhena-gorgias-policy-resolution-2026-09-21 (712b1e5753cd9b871881d328d79ecb20ccbb279c045fe9233891c6805441133b). Read each source method before interpreting differences.

Frozen method SHA-256 712b1e5753cd9b871881d328d79ecb20ccbb279c045fe9233891c6805441133b

Read every conversation behind these scores

Detailed decisions, source records and available study downloads require a verified work email. Your email and view or download activity are shared privately with Alhena. Privacy details.

Considering Alhena? See how its shopping and support agents would handle your customers’ questions.

Book a demo