CTGT's New AI Platform Enhances DeepSeek Model Accuracy
CTGT's upgraded platform removes bias and hallucinations from AI models like DeepSeek, enabling them to answer sensitive questions with 96% accuracy. The solution was awarded the top prize at VentureBeat Innovation Showcase 2025.

In testing, CTGT enabled DeepSeek to answer 96 percent of sensitive questions, compared to 32 percent originally. CTGT was named winner of VentureBeat Innovation Showcase 2025. CTGT announced an upgrade to its platform that removes bias, hallucinations, and other unwanted model features from DeepSeek and other open source AI models. CTGT demonstrated its new platform upgrades at VentureBeat Transform 2025 and was named the top winner of the Showcase. CTGT has built infrastructure for a leading insurance provider and is working with a major bank on email compliance. McKinsey reported $67.4 billion in global losses linked to AI hallucinations in 2024. Deloitte data shows that 47 percent of enterprise AI users made major business decisions based on hallucinated output. CTGT launched in February 2025 with funding from Gradient, General Catalyst, Y Combinator, and notable angels. CTGT enables enterprises to deploy AI for high-risk use cases by removing bias, hallucinations, and censorship automatically in real time.
According to the source: The Manila Times.
What's Your Reaction?






