Muteki Group

Muteki Group was hired by RESQUAD AI, a staff augmentation company, to test a client's solution. The project involved a detailed comparison of the performance of a deep learning model for predicting stock price movements using Limit Order Book data before and after implementing Intelligent Storage Virtualization. The primary goal was to evaluate the impact of this storage virtualization solution on various performance metrics, such as data access speed, computational resource usage, and scalability, compared to traditional methods using AWS S3 and Boto3. The project was successful, with Muteki Group's expertise and professionalism ensuring a thorough evaluation of the solution’s functionality and reliability.

In a project for an IT services company, Muteki Group was tasked with building the backend and AI-agent layer for an enterprise assistant platform. The main goals included developing scalable APIs and infrastructure to support AI workflows, implementing RAG-based agents using LLMs, FAISS, and Qdrant for semantic search, orchestrating multi-step agent logic via LangChain and LangGraph, and integrating with PostgreSQL, Neo4j, and external document sources. Muteki Group successfully delivered a robust foundation for the company's intelligent assistant system.

Kitrum, a software development firm, hired Muteki Group to design and deliver the backend, AI/ML pipelines, and DevOps infrastructure for their MVP project in the fashion domain. The project involved designing and implementing the architecture and APIs to support user onboarding, wardrobe digitization, and personalized outfit recommendations. Muteki Group also built the full image-processing pipeline for garment recognition, feature extraction, and integration with large language models to generate personalized style suggestions. On the DevOps side, they delivered a complete cloud infrastructure setup with automated deployments, monitoring, and production-ready operations. The project was successful, resulting in a production-ready MVP backend capable of scaling for future feature expansion.