Backend Developer (AI / LLM)

Company Description MindHive | AI Workforce Automation, AI Forecasting and AI Chatbots designs tailor-made AI applications and infrastructure that help businesses achieve measurable results and competitive advantage. The team of AI specialists delivers bespoke solutions across domains such as machine learning, data analytics, digital transformation, and AI-driven business applications. MindHive’s work spans optimizing operations, improving customer experiences, and enabling organizations to harness data for strategic decision-making. As active thought leaders, the company regularly shares insights on emerging AI trends, keeping clients informed and prepared for the future of work. About The RoleWe're building an AI-powered platform that turns messy, real-world business input into structured, reliable output. Two sides to the work: an AI retrieval and LLM pipeline that does the intelligent matching, and an integration layer that keeps our platform in two-way sync with the external systems our customers already use. We're looking for a backend developer who's comfortable across both. What You'll Do· Build and maintain the backend services at the core of the platform. · Work on our AI retrieval and matching pipeline — combining search and LLMs to produce accurate, structured results. · Integrate LLMs into production workflows — reliable prompting, structured output, and keeping cost and latency in check. · Design and run evaluations — build benchmarks, measure quality, and catch regressions before they ship. · Build and maintain two-way integrations with third-party business systems — keeping data in sync reliably and correctly. · Own the hard parts of sync: reliability, retries, conflict handling, and clean identity mapping across systems. · Build onboarding and data pipelines for bringing new customers onto the platform. · Write clean, well-tested code and collaborate across the team. Must-Have· 3+ years building production Python backends — services, APIs, data pipelines. · Integration experience — you've built system-to-system sync or third-party API work and handled retries, idempotency, webhooks, and schema drift. · Hands-on LLM experience — you've shipped something real with an LLM, not just prompted a chatbot. · Understands evaluation — knows why it matters, can build a benchmark, and reasons clearly about quality trade-offs. Non-negotiable. · Strong SQL and data modeling; comfortable with messy real-world data. · Solid testing discipline and git workflow. Nice-to-Have· Experience with vector search / RAG and hybrid retrieval. · Background in ERP, accounting, or POS domains. · Cloud and infrastructure-as-code experience. · Event-driven or adapter-pattern architecture experience. Tech Stack· Language: Python · Backend: REST APIs, async services, task queues · Data: Relational databases, vector search, ETL at scale · AI: LLM APIs, embeddings, retrieval pipelines, evaluation harnesses · Infra: Containerized services, cloud, infrastructure-as-code