TICKR

TICKR AI models significantly outperform leading frontier models from OpenAI and Anthropic in dynamic hierarchical product categorization—a pervasive pain point we solve for our CPG and retail clients. Leveraging deep domain and technical expertise, our LLMs achieve up to 98.1% accuracy, surpassing GPT-4o and Claude Sonnet 3.5 under the same instructions (prompts) and experimental conditions. Read our latest blog post to learn how we're setting new standards in AI for CPG and retail. https://lnkd.in/gh9kAspy

Samuel K., our VP of AI & Data Science, recently joined CorrDyn to discuss the frontier of AI in CPG & Retail. They delve into the unique challenges of this space and how TICKR is building and deploying AI solutions that harness the Cambrian Explosion of emergent use cases, now realistic to solve with AI and AI-assisted data science. Have a listen!

Discover the work our AI & Data Science team has done using Shopper Journey Data to construct and analyze walk-out rates. Using precise timestamps and basket-level data, we can track whether shoppers made additional shopping trips for specific products in the same day. Our findings reveal substantial variation in walk-out rates across different retailers, food categories, and brands. These differences may be influenced by product assortments, pricing, and promotion strategies. These insights provide our clients with actionable steps to reduce walk-out rates. We're excited about the impact this will have on our clients and partners! https://lnkd.in/gA23gGzm

Check out the work our AI & Data Science team has done identifying substitutable products across retailers and geographic regions. By analyzing extensive shopping cart data and using sequential probabilistic models, we can uncover how the ranking of substitutes varies across retailers and geographic regions. These insights help our clients understand the substitutes of their products based on consumer purchase behavior. We are excited about the impact this will have on our clients and partners! https://lnkd.in/gQDy7-uv

Check out the work our AI & Data Science team has done modeling consumer preferences using qualitative survey data and large language models! This greatly improves and enriches the consumer survey process and research by augmenting known sentiment, extending surveys beyond the original question set at a rapid pace, providing a far more flexible framework for working with qualitative data from real respondents and their 'digital twins', and much more! Link: https://lnkd.in/g4A8JmAV

https://lnkd.in/gJYsTcRn Excited to share that our AI & Data Science team has developed a system called Generative Predictor Search that leverages large language models (LLMs) to enhance our forecasting models. By utilizing the parameterized knowledge of LLMs, we identify external variables, from data source like the FRED or news, that significantly improve our multivariate forecasts. The results are impressive: - Reduced error rates in our models - Elimination of spurious predictors - Performance that is an order of magnitude faster than alternatives search methods We're looking forward to the impact this will have on our clients & partners!

See how an AI assistant (that learns from interacting with you) can unlock analytics at your organization. https://lnkd.in/gR56E5ru

AI powered assistants can help solve your analytics bottleneck: https://lnkd.in/gN_UTihw

https://lnkd.in/gzQSzsZE Check out the latest blog post from our product team about Tickr's Explainable Forecasting (XF) solution. XF combines Tickr's cutting-edge forecasting stack with our proprietary LLM Predictor Search, democratizing business planning and forecasting at scale. Stay tuned for Parts 2 and 3, which will dive into the technical aspects (LLM Search, RAG, Forecasting) of our XF solution.

https://lnkd.in/gzQSzsZE Check out the latest blog post from our product team about Tickr's Explainable Forecasting (XF) solution. XF combines Tickr's cutting-edge forecasting stack with our proprietary LLM Predictor Search, democratizing business planning and forecasting at scale. Stay tuned for Parts 2 and 3, which will dive into the technical aspects (LLM Search, RAG, Forecasting) of our XF solution.