Loading
GeekRabit is an applied-AI engineering studio. We take GenAI features from pilot to production inside the Java and Spring Boot platforms companies already run, on AWS. Most GenAI pilots stall on the same things: integration with existing services, data pipelines, access control, evaluation, and cost per request. That is the work we do. What we do - Retrieval-augmented generation over enterprise documents on Amazon Bedrock, with retrieval quality measured against a test set - LLM capabilities added to existing Spring Boot services without a rewrite - Production readiness for AI features: guardrails, evaluation pipelines, observability, prompt caching and cost control - Conversational workflows on WhatsApp and chat: multilingual intent extraction, confirmation loops, human-in-the-loop fallbacks - End-to-end product delivery: Spring Boot backends, React web apps, Flutter mobile apps, PostgreSQL, AWS infrastructure We work with Amazon Bedrock, Anthropic Claude, OpenAI GPT and open-source models, and choose per use case. Proof of work: Thiya, a commerce platform for recurring local services, designed and built end to end. Its WhatsApp layer reads mixed Hindi and English customer messages, resolves intent with an LLM, confirms with the customer and updates orders. Led by a senior engineer with 12+ years building systems for global banks, payment processors and trading platforms. AWS Certified Generative AI Developer - Professional and Solutions Architect - Associate. We partner wit