ScriptAI for Codev1.1

Recommendation Engine Setup

by Community · open-source · Last verified 2026-03-17

Builds a two-stage recommendation engine using a two-tower neural retrieval model for candidate generation and an LLM-based cross-encoder for re-ranking, with a Feast feature store for real-time user context. Supports cold-start via content-based fallback and A/B test scaffolding for algorithm experimentation.

https://github.com/NVIDIA-Merlin/Merlin ↗

C+

C+—Average

Adoption: BQuality: AFreshness: ACitations: C+Engagement: F

Specifications

License: MIT
Pricing: open-source
Capabilities: two-tower-retrieval, llm-reranking, cold-start-fallback, ab-test-support
Integrations: pytorch, feast, redis, faiss, fastapi
Use Cases: e-commerce-recommendations, content-discovery, job-matching
API Available: Yes
Language: python
Dependencies: torch, feast, faiss-cpu, fastapi, redis, numpy
Environment: Python 3.10+, CUDA optional
Est. Runtime: Training: 1-4 hours; inference: <20ms
Tags: recommendation, collaborative-filtering, llm-reranking, two-tower, personalization
Added: 2026-03-17
Completeness: 100%

Index Score

58.7

Adoption

Quality

Freshness

Citations

Engagement

Explore the full AI ecosystem on Agents as a Service