Post-train a small specialist model for financial question answering to pair with a frontier orchestrator
Instead of serving every request with an expensive, slow frontier model, train a smaller open-weight model on the specific question-answering subtask (e.g. retrieving and reasoning over dropped-in spreadsheets) and hand off work from a frontier orchestrator to it. Ramp Labs post-trained a 35B Qwen model that beat the frontier model's exact task accuracy while serving at Haiku-level latency.
online B2B Deep Tech / Infrastructure
From AI Engineer — Our Smaller Model Beat Opus on Finance Tasks - Alexander Shevchenko, Ramp at 03:06
Problem: Frontier models are slow and expensive on thinking tokens; much of an agent's time is spent on a subtask (understanding spreadsheets, answering domain questions) that a smaller specialist model can do as well or better, at far lower latency and cost.
For: Finance and accounting teams using agentic spreadsheet/analysis products, or any vertical product where a narrow, high-frequency reasoning subtask dominates latency and cost.
Products from this video
Examples
- Ramp Sheets: Ramp's spreadsheet agent, launched November before frontier labs shipped their own spreadsheet agents, whose backend gained the post-trained specialist model
Other takes on AI model development
- Retrain small open-source language models for long-horizon agentic workloads and sell efficient inference and on-device deployments
- AI simulation company that models human behavior and lets organizations test decisions against simulated populations
- An open foundation-model company focused on coding and long-horizon software tasks
- Enterprise context and training-data infrastructure for AI applications
- In-database semantic analytics using large database models
- A domain-specific AI product built around private data, specialized models, and purpose-built workflows