Xinobi AI Tech Stack

Personal AI agents with proprietary memory and tool orchestration

Software Development Tokyo 11–50 employees Founded 2024 Privately Held

Xinobi AI is a Tokyo-based AI startup (founded 2024) building personal AI agents through a stack centered on LangChain, LangGraph, and vector databases (Faiss, Milvus, Weaviate, Pinecone). The tech choices—RAG pipelines, LLM evaluation frameworks, and multi-cloud deployment (GCP, AWS, Azure)—signal a focus on retrieval-augmented agent capabilities rather than foundational model work. Hiring is balanced across engineering, product, and marketing with manager-level roles present, reflecting early-stage scaling beyond pure R&D into go-to-market execution.

Tech Stack 32 technologies

Core StackLooker Amplitude Python FastAPI LangChain Django Weaviate Pinecone OpenAI Anthropic Hugging Face Docker Kubernetes Weights & Biases MLflow AWS BigQuery Google Analytics Figma Google Analytics 4 LangGraph Faiss Milvus Google Mistral Git BentoML GCP Azure SQL+2 more

What Xinobi AI Is Building

◆Challenges

Cross-border coordination for product launch
Expanding personal ai to new markets
Localizing product for korean market
Scaling personal ai from zero to ten
Defining new product category
Scaling adoption globally
Building brand consistency across regions
Cross-border execution readiness
International project coordination

▲Active Projects

2025 product launch across japan, korea, u.s.
International r&d / poc / partnership projects
Personal ai agents backend system
Global go-to-market strategy
Community & ecosystem development
Memory and retrieval pipelines for rag
Benchmarking and evaluation frameworks for llms
Personal ai product launch 2025
Gtm readiness for japan, korea, u.s.
0→1 and 1→10 phases of personal ai

Hiring Activity

Steady7 roles · 3 in 30d

Department

Engineering

Marketing

Product

Seniority

Mid

Manager

Senior

Company intelligence

Find more companies like Xinobi AI by tech stack, pain points and active projects

Get started free

About Xinobi AI

Xinobi AI develops personal AI agents designed to think, learn, and grow alongside users. The company operates across Tokyo, Seoul, and Singapore, with a co-founding team that includes former Silicon Valley executives and a serial entrepreneur. The product roadmap centers on 2025 launches in Japan, Korea, and the U.S., supported by ongoing work on RAG memory systems, LLM benchmarking, and global GTM strategy. Current hiring focuses on expanding the core engineering and product teams while building out marketing capacity for international expansion.

HeadquartersTokyo

Company Size11–50 employees

Founded2024

Hiring MarketsJapan

Frequently Asked Questions

What tech stack does Xinobi AI use?

LangChain, LangGraph, FastAPI, Python, and vector databases (Faiss, Milvus, Weaviate, Pinecone). For model access: OpenAI, Anthropic, Mistral, Google, Hugging Face. Infrastructure: GCP, AWS, Azure, Kubernetes, Docker. Observability: Weights & Biases, MLflow, BentoML.

What is Xinobi AI working on in 2025?

Personal AI product launch across Japan, Korea, and the U.S.; memory and retrieval pipelines for RAG; LLM evaluation frameworks; global GTM strategy; and market localization (particularly Korea). Also scaling agent backend infrastructure and building cross-border execution capability.