For specific, high-leverage tasks like conversation summarization and re-ranking search results, Intercom trains its own custom models. These smaller, fine-tuned models have proven to be cheaper, faster, and higher quality than using general-purpose frontier models from vendors.
Related Insights
The top 1% of AI companies making significant revenue don't rely on popular frameworks like Langchain. They gain more control and performance by using small, direct LLM calls for specific application parts. This avoids the black-box abstractions of frameworks, which are more common among the other 99% of builders.
Simply offering the latest model is no longer a competitive advantage. True value is created in the system built around the model—the system prompts, tools, and overall scaffolding. This 'harness' is what optimizes a model's performance for specific tasks and delivers a superior user experience.
For specialized, high-stakes tasks like insurance underwriting, enterprises will favor smaller, on-prem models fine-tuned on proprietary data. These models can be faster, more accurate, and more secure than general-purpose frontier models, creating a lasting market for custom AI solutions.
The vast majority of Intercom Fin's resolution rate increase came from optimizing retrieval, re-ranking, and prompting. GPT-4 was already intelligent enough for the task; the real gains were unlocked by improving the surrounding architecture, not waiting for better foundation models.
Instead of a generalist AI, LinkedIn built a suite of specialized internal agents for tasks like trust reviews, growth analysis, and user research. These agents are trained on LinkedIn's unique historical data and playbooks, providing critiques and insights impossible for external tools.
The primary driver for fine-tuning isn't cost but necessity. When applications like real-time voice demand low latency, developers are forced to use smaller models. These models often lack quality for specific tasks, making fine-tuning a necessary step to achieve production-level performance.
The "agentic revolution" will be powered by small, specialized models. Businesses and public sector agencies don't need a cloud-based AI that can do 1,000 tasks; they need an on-premise model fine-tuned for 10-20 specific use cases, driven by cost, privacy, and control requirements.
Building a single, all-purpose AI is like hiring one person for every company role. To maximize accuracy and creativity, build multiple custom GPTs, each trained for a specific function like copywriting or operations, and have them collaborate.
Basic supervised fine-tuning (SFT) only adjusts a model's style. The real unlock for enterprises is reinforcement fine-tuning (RFT), which leverages proprietary datasets to create state-of-the-art models for specific, high-value tasks, moving beyond mere 'tone improvements.'
Instead of offering a model selector, creating a proprietary, branded model allows a company to chain different specialized models for various sub-tasks (e.g., search, generation). This not only improves overall performance but also provides business independence from the pricing and launch cycles of a single frontier model lab.