Super-App Recommendation and Personalisation

LLM-powered enrichment plus hierarchical clustering.

0in Indonesia

Named Models

Wide and DeepGoogleReleasedJun 2016Runs onnot in the notesClassrecommendationRecommender architecture that trains a wide linear model and a deep network together, productionised on Google Play. Paper by Cheng et al. (2016).Commercial use: not stated; Jakarta hosting: unknown; Indonesian support: not stated

Where this comes from

Models the research names for this use case. Research, with sources on each model record.

Who Runs It

production, 1

GrabCountry: not in the notesProperty-data personalisation (LLM enrichment plus hierarchical clustering)Vendor: GrabproductionDate6 Oct 2026Evidence: company statementConfidence: lowSource: engineering.grab.com

Where this comes from

Documented uses by named organisations. The evidence column says who made the claim.

Candidate Models

SEA-HELM leaderboard, 58 models, SEA average

A general benchmark for this model class, not a test on this use case. SEA average and per-language (incl. Indonesian) scores for open-weight models. Higher is better. Showing 10 of 11.

Qwen3.6-35B-A3BAlibaba (Qwen) maintainer-run73.86

Qwen3.5-35B-A3BAlibaba (Qwen) maintainer-run73.55

gemma-3-27b-itGoogle maintainer-run64.54

Gemma-SEA-LION-v4-27B-ITAI Singapore maintainer-run64.35

gpt-oss-20bOpenAI maintainer-run60.64

Llama-3.3-70B-InstructMeta maintainer-run60.44

Qwen3-8BAlibaba (Qwen) maintainer-run60.33

GLM 4.7 FlashZ.ai (Zhipu) maintainer-run53.76

Gemma-3-4B-itGoogle maintainer-run47.84

Nemotron 3 Nano 30B-A3BNVIDIA maintainer-run46.89

050100

Nemotron-SEA-LION-v4.8-120B-A12BAI Singapore, LLMReleasedSep 2026open weightsIndonesian statedhas scoreslast 18 months

Nemotron-SEA-LION-v4.8-30B-A3BAI Singapore, LLMReleasedSep 2026open weightsIndonesian statedhas scoreslast 18 months

Qwen-SEA-LION-v4.5-27B-ITAI Singapore, LLMReleasedMay 2026open weightsIndonesian statedhas scoreslast 18 months

Qwen3.6-35B-A3BAlibaba (Qwen), LLMReleasedApr 2026open weightsruns on edgehas scoreslast 18 months

Nemotron 3 SuperNVIDIA, LLMReleased2026open weightsIndonesian statedhas scoreslast 18 months

Nemotron 3.5 Lightning 30B-A3BNVIDIA, LLMReleased2026open weightsIndonesian statedhas scoreslast 18 months

gpt-oss-120bOpenAI, LLMReleasedAug 2025open weightsIndonesian statedhas scoreslast 18 months

Granite 4.2 8BIBM, LLMReleasedSep 2026open weightshas scoreslast 18 months

Hy4-previewTencent, LLMReleasedAug 2026open weightshas scoreslast 18 months

GLM-5.3Z.ai (Zhipu), LLMReleasedAug 2026open weightshas scoreslast 18 months

Showing 10 of 411. All models of this class

Where this comes from

Derived, not researched: models in the directory whose class matches a class this use case uses, excluding family records and models shut down or deprecated. Ranked by how many fit facts they meet, then by release date. Each fit fact is read from the model record.

Use Case

Industry: retail and logistics
Model classes: LLMrecommendation
Benchmarks or datasets: not in the notes
Regulatory notes: Surat Edaran Menkominfo 9/2023 Etika Kecerdasan Artifisial, Kementerian Komunikasi dan Informatika (now Kementerian Komunikasi dan Digital)Super-app recommendation and personalisation is an AI-based service run by an electronic system operator, and the circular asks such operators to build and use AI on the values of inclusivity, transparency, humanity and security.
UU 27/2022 Pelindungan Data Pribadi, DPR and President of the Republic of Indonesia (administered by Kementerian Komunikasi dan Digital)Super-app recommendation and personalisation processes personal data of shoppers, sellers, drivers or customers, so the controller needs a lawful basis, must protect the data and must honour data subject rights.
Permenkominfo 5/2020 Penyelenggara Sistem Elektronik Lingkup Privat, Kementerian Komunikasi dan Informatika (now Kementerian Komunikasi dan Digital)Super-app recommendation and personalisation runs on an online platform used by people in Indonesia, so the private electronic system operator must register and meet this regulation's duties on content and data.

Where this comes from

From the industry tracks T08 to T10 of the research notes.

R&D Topics

Links to R&D topics are awaiting review.

Where this comes from

Which R&D topics would most change this use case. Drafted by Claude from the topic definitions and the use case; proposed until approved.

Incidents

Amazon Reportedly Sold Products and Recommended Frequently Bought Together Items That Aid Suicide Attempts2022-02-04, safetyDespite complaints notifying Amazon about the sale of various products that had been used to aid suicide attempts, its recommendation system reportedly continued selling them and suggesting their frequently bought-together items.
Amazon Reportedly Recommends Explosive-Producing Ingredients as 'Frequently Bought Together' Items for Chemicals2017-09-18, safetyAmazon was reported to have shown chemical combinations for producing explosives and incendiary devices as "frequently bought together" items via automated recommendation.

Where this comes from

Documented failures, harms or misuse in this kind of use, Indonesia first, then the region, then well-known cases elsewhere (AI Incident Database, OECD AI incidents monitor, reputable news). Not a claim about any deployment listed here.

Opportunities

Opportunities are awaiting review.

Where this comes from

Gaps found by rule in what the directory holds for this use case. Each is a proposal for review.

Sources

Grab Tech engineering.grab.com https://engineering.grab.com/

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Super-App Recommendation and Personalisation", https://aurion.semai.id/usecases/ret-recommendation-personalisation/, records of October 2026.

Super-App Recommendation and Personalisation

Named Models

Who Runs It

Candidate Models

SEA-HELM leaderboard, 58 models, SEA average

Links

Models 1

Deployments 1

Sources