Stampli uses Codex to accelerate Deep Finance launch production
Stampli, an intelligent procure-to-pay platform, used OpenAI’s Codex and ChatGPT Work to support the launch of its Deep Finance product. The company estimates that defined go-to-market and content-production work fell from about 243 modeled active role-hours to approximately 77 hours, a reduction of roughly 166 hours and a 3.16x faster production cycle.
Deep Finance turns data moving through Stampli’s procure-to-pay platform into executive spend intelligence for CFOs, VPs and other business leaders. Stampli moved the offering from an initial prototype demonstration to a public go-to-market launch and first shipped product in about six weeks.
A constrained launch workflow
The launch required product development, positioning, design, communications, enablement and operations to move in parallel. Stampli said its design resources and outside contractors had already been allocated to other work, so its marketing team used Codex to bring product context, meeting notes, decisions and messaging guidance into a shared system.
That system helped the team convert changing product decisions into assets ready for review. The launch included a seven-part blog series, emails, a webinar and supporting deck, social and paid creative, a PR Newswire release, a Deep Finance web page and sales enablement materials.
Codex also supported the hero animation through exploration, iteration and packaging. Stampli estimates that it handled roughly 90% of the polished animation work, before a contractor completed the opening scene and final format. The company said that all customer-facing materials retained human review and final approval.
From launch support to daily product marketing
The infrastructure used for Deep Finance is also part of Stampli’s everyday product-marketing workflow. Previously, keeping materials current meant interviewing product managers, reading Jira tickets, reviewing GitHub activity and working through meeting notes before translating the findings into help-centre articles, presentations and one-pagers.
Stampli has automated much of that work with a GPT-powered system that gathers information from product systems and meeting notes, then helps maintain current materials. Agents connected to the company’s source of truth can also produce material for its website and social channels. Director of Product Marketing Melad Zahedi said this expanded a small team’s output to hundreds of pieces of content each week, compared with only a couple before.
The operational pattern resembles Australian Payments Plus AI workflow acceleration in its use of ChatGPT Enterprise and Codex to streamline work, while Stampli’s case centres on connecting product knowledge to launch production. In an executive meeting, Zahedi said an employee used Codex to retrieve and analyse metrics from HubSpot and other systems during the call, replacing work that would have taken the FP&A team half a day to prepare.
Implications for business teams
Stampli reports that teams in product, marketing, customer success, sales and enablement now use ChatGPT Work and Codex from prototype through launch. The time saved has allowed product marketing to spend less time reconstructing context and more time advising leaders on product and company strategy.
For businesses, the case points to a practical approach: connect approved internal context to a defined workflow, keep human review for external outputs, and assess results through measured active work and delivery time rather than AI use alone.