Posts by Sarah Chen — Harbor Software

The MVP Trap: When Minimum Viable Becomes Maximum Technical Debt

The MVP concept is one of the most misunderstood ideas in product development. Eric Ries defined it as the version of a product that allows a team to collect the maximum amount of validated learning with the least effort. Somewhere between his book and actual

Structured Output from LLMs: Reliable JSON Every Time

Getting an LLM to return valid JSON sounds trivial. Ask for JSON, get JSON. In practice, LLMs produce invalid JSON at a rate between 2% and 15% depending on the model, prompt complexity, and output schema. For a feature that makes 10,000 API calls per

OpenAI vs Anthropic vs Open Source: Choosing Your LLM Provider

Choosing an LLM provider is one of the most consequential technical decisions you will make in 2023. It affects your cost structure, latency profile, compliance posture, and feature velocity for years to come. It is also a decision that most teams make badly, either by

Building AI Agents That Actually Work

The AI agent hype cycle is in full swing. Every week a new framework promises autonomous agents that can browse the web, write code, manage your calendar, and negotiate your rent. The demos are impressive. The production reality is different. Most agent systems are fragile,

Prompt Engineering for Production Applications

Most prompt engineering advice comes from people tinkering with ChatGPT. They share tips like “be specific” and “give examples” as though these insights are revelatory. Building prompts for production applications is a fundamentally different discipline. You are not crafting a single clever instruction; you are

Natural Language Processing for Enterprise Search

Enterprise search is broken in most organizations, and almost everyone has accepted the brokenness as normal. The search bar exists on the company intranet, the document management system, the knowledge base, the wiki, the ticketing system, and a dozen other tools. Users type a query,

Monitoring and Alerting for AI-Powered Applications

Traditional monitoring assumes deterministic systems. You send the same input, you get the same output, and any deviation is a bug worth investigating. AI-powered applications break this assumption completely. A language model might return different text for the same prompt on consecutive calls. A recommendation

Understanding Embeddings: From Theory to Production

Embeddings are the bridge between human-readable data and machine-processable mathematics. They are the input to vector search, the foundation of recommendation systems, and the representation layer in modern NLP. Despite their centrality, most engineering teams treat embeddings as a black box: call an API, get

Why We Chose Python and FastAPI for Our AI Backend

When we started building Harbor Software’s core inference platform in late 2021, the backend framework decision felt unusually consequential. We were building something that needed to serve ML model predictions with sub-200ms latency, handle concurrent long-running inference jobs, and remain approachable enough that a team

Author: Sarah Chen