Interest-graph social network for Gen Z with gamified features and AI companionship
Soul App is a social platform designed around interest-based discovery and gamified mechanics, targeting China's Gen Z audience. The tech stack reveals heavy investment in ML infrastructure (TensorFlow, PyTorch, GPT, Qwen, LLaMA, DeepSpeed) paired with big-data processing (Hadoop, Spark, Hive), suggesting the company is scaling beyond social feeds into personalization and emerging AI features. Active projects around AI emotional companionship, digital humans, and speech synthesis indicate a pivot toward conversational AI experiences — moving beyond traditional social networking.
Soul App launched in 2016 as an interest-graph social network with gamified product design for younger users in China. The platform combines group chat parties, gifting mechanics, and low-latency social experiences to build what the company frames as a social metaverse. The company operates from Shanghai with 1,001–5,000 employees and is currently hiring across product, engineering, data, and marketing, though hiring velocity has slowed. Active projects span recommendation optimization, content engagement, monetization efficiency, and emerging AI-driven features like emotional companionship and digital dialogue.
C++, Java, Python, Hadoop, Hive, Apache Spark, TensorFlow, PyTorch, GPT, Qwen, LLaMA, DeepSpeed, Go, Docker, and SQL. Also uses design tools (Adobe Premiere Pro, Photoshop, Illustrator) and major LLM models (GPT, ChatGPT, Qwen, LLaMA).
Current projects include recommendation ranking optimization, AI emotional companionship applications, digital human dialogue deployment, speech synthesis, product commercialization, and content engagement improvement. Also prioritizing monetization efficiency and algorithm optimization.
Other companies in the same industry, closest in size