Video: Shaped @ 2024 Year End Gen AI Zoo

An interview between OSS4AI CEO Yujian Tang, and Shaped's Co-Founder and CEO, Tullie Murrell at the 2024 Year End Gen AI Zoo. In this video Tullie explains how Shaped simplifies AI-powered search, recommendations, and personalization by providing an end-to-end platform that integrates data ingestion, fine-tuning, and real-time re-ranking. Unlike vector databases, Shaped focuses on delivering highly relevant results by combining semantic search, keyword methods, and session-based re-ranking. The platform allows businesses to easily customize optimization goals and use real-time user interactions to improve relevance without frequent model retraining. Tully highlights Shaped’s role in powering applications like e-commerce recommendations, personalized RAG systems, and emerging conversational recommenders.

What is Shaped?

Shaped is an end-to-end AI platform designed to power personalized search, recommendations, and marketing content. Unlike traditional vector databases that focus on storage and retrieval, Shaped provides a complete solution, combining data ingestion, real-time re-ranking, and model optimization. Businesses can quickly achieve relevance for their content, products, or user experiences while saving time and effort.

Why It Matters

Tullie highlighted a fundamental shift: consumers today expect hyper-relevant content as a baseline experience. Whether it’s a personalized feed, search results, or product recommendations, relevance drives both user satisfaction and business revenue. Drawing from his experience at Meta, where he worked on ranking systems for Instagram and ads, Tullie emphasized that small improvements in relevance can yield substantial commercial impact.

How Shaped Stands Out

Shaped is built to handle the entire relevance problem, not just data storage:

  1. End-to-End Solution: Shaped combines semantic search, keyword retrieval, and real-time re-ranking to deliver the most relevant results.
  2. Real-Time Adaptability: By ingesting real-time event streams (e.g., clicks, purchases), Shaped dynamically adjusts recommendations without requiring frequent model retraining.
  3. Customizable Goals: Businesses can define objectives like conversions, engagement, or diversity through an intuitive interface, enabling fine-grained optimization.
  4. Personalized RAG and Conversational AI: Shaped also supports advanced use cases like retrieval-augmented generation (RAG) and conversational recommenders, a growing area in AI-powered discovery.

Key Takeaways

Shaped’s focus is clear: help businesses deliver exceptional, relevant user experiences through easy-to-implement, real-time AI solutions. By bridging the gap between personalization and machine learning infrastructure, Shaped empowers teams to compete with tech giants without needing extensive resources.

For those eager to explore, Shaped offers self-serve demos showcasing its capabilities for movie recommendations, e-commerce, and more. As the future of digital discovery evolves, platforms like Shaped will play a pivotal role in powering seamless, AI-driven user experiences.

Get up and running with one engineer in one sprint

Guaranteed lift within your first 30 days or your money back

100M+
Users and items
1000+
Queries per second
1B+
Requests

Related Posts

Tullie Murrell
 | 
August 31, 2022

A ranking model for every use-case!

Tullie Murrell
 | 
March 22, 2023

Shaped 1.0: The fastest way to personalize your product, platform or marketplace

Daniel Camilleri
 | 
March 29, 2022

YC Demo Day W22