Entry Point

Entry Point is a platform for turning business data into task specific AI models. It combines three tools in one workflow: Transforms, which run a prompt across every row of a dataset to tag, extract or classify; Synthetic Data, which expands a handful of labeled examples into a full training set that a reviewer approves or rejects; and Fine tuning, which trains a custom model and compares it against the original prompt. It is aimed at teams building an AI feature or automating an internal process, such as support ticket prioritization, data extraction, content moderation, fraud detection, data enrichment and recommendations. The workflow runs from writing a prompt and running it at scale through capturing completions, quality checks, fine tuning across providers and pointing an API at the resulting model. Data import, prompt and completion templates, token counts and cost estimates, a playground, and CSV and JSONL export are included on every plan. Pricing is published and tiered by rows. Starter costs $49 per month for up to 5,000 rows and 3 seats, Growth costs $99 per month for up to 25,000 rows and 5 seats, and Pro costs $249 per month for up to 100,000 rows and 10 seats with premium support. Every plan includes the full suite, and an Enterprise option is available by contacting the vendor. A free start allows up to 300 rows. The row limit is pooled across projects, and fine tuning and data generation run on the customer's own connected provider accounts, such as OpenAI, so provider costs are billed separately.