Optimizely has purpose-built marketing models trained specifically for marketing tasks, achieving 10x better cost efficiency than frontier models with ~7% better quality.
Public source
Publisher name
Public post
Some great insights coming out of Opticon NYC. Looking forward to what Opticon London brings in October. 1. Opti is post-training its own models for marketing. They're b…
Company
Optimizely
- Industry
- Software Development
- Location
- New York, US
- Company size
- 1,001–5,000 employees
About Optimizely
Optimizely is the AI platform for marketing, built for the experience-makers who shape how brands show up in the world. Bringing together content management, content marketing, experimentation, commerce, personalization, and analytics into a single platform, Optimizely gives marketers the freedom to create, customers the experiences they remember, and AI agents the content they come looking for. With Optimizely, you're free to grow. Learn more at optimizely.com.
See moreLatest activity
Latest activity from Optimizely
76 signals
Presence & Recognition
Optimizely is hosting a salon event in Barcelona on September 22 featuring a business council session led by Kasia Odrozek.
Presence & Recognition
Optimizely completed the first Opticon in NYC.
Presence & Recognition
Optimizely is hosting a conversation on Tuesday September 15th for Women of Customer Success hosted by Synthesia on the topic of money, quotas, and trust.
Discover more
Similar signals
Similar public activity from other companies.
Research & Knowledge
Jasper
Jasper published the Bloom Technical Report, a complete account of the year of work on building a production-grade image model from scratch, live on Hugging Face.
Research & Knowledge
Enterspeed
Enterspeed ran AI enrichment on real retail data to achieve a 20% increase in traffic, 40% increase in time on site, and a 150+ like-for-like revenue index.
Research & Knowledge
Optivalue.ai
Optivalue.ai has been cited by Gartner in a new research note titled From Adoption to Impact: The New AI-Driven Customer Marketing Equation, published on 3 August 2026, referencing Optivalue.ai as an example of a platform embedding proof of value realization directly into the customer experience.
Research & Knowledge
Magic
Magic developed a new pretraining recipe for Frontier that matches DeepSeek V4 Pro's pretrain using 50x less compute, roughly half the FLOPs used for GPT3, or ~$0.5M on GB200.
Research & Knowledge
Perplexity