Making Sense, a technology consulting firm that builds software for mid-market and private equity-backed companies, has been named a Select partner in the Services Track of Anthropic’s Claude Partner Network.
The designation is based on AI products the firm has taken live for clients.
Following two years’ worth of pilots that rarely made it into production, companies hiring outside help for AI have been getting pickier; buyers want evidence that a firm has shipped working systems before they sign.
Anthropic launched the Services Track in June to rank the firms that help companies build and deploy with Claude. Partners are measured on three things: certified practitioners on staff, customers deployed in the trailing 12 months, and public customer stories.
Making Sense has 20 people certified through the Anthropic Partner Academy and now appears in the public Claude Partner Hub directory. Its case rests on two clients: Doppler, a Latin American marketing platform, and Viallion, an AI-powered trading automation company.
Why buyers want proof
For startups and scaleups, the cost of hiring the wrong AI partner is high. A stalled engagement can use up months of runway, and smaller teams rarely have in-house specialists who can audit a consultancy’s claims.
The record so far certainly gives buyers a reason to be careful. An MIT NANDA study published last year found that 95% of generative AI pilots had produced little to no measurable impact on profit and loss, and that companies building on their own failed more often than those working with outside partners.
Vendor claims haven’t made the choice easier. Gartner has warned about agent washing, whereby existing chatbots and automation tools are rebranded as agentic AI. The firm estimated, in fact, that only about 130 of the thousands of vendors marketing agentic products offer the real thing.
On the supply side, Anthropic’s own partner network drew more than 40,000 applications within months of opening.
The Services Track is Anthropic’s answer to that sorting problem; instead of relying on a firm’s own marketing, it publishes thresholds tied to delivery. Under Anthropic’s published requirements, a Select partner needs at least 10 certified practitioners, two or more customers taken live with Claude in the past year, and at least one public customer story.
Certifications belong to individuals and are earned through exams, so the count reflects people a firm actually employs.
“Anthropic reviews each partner’s standing every quarter, based on its certified professionals and the customers it has taken live with Claude,” said Fernando Florez, chief revenue officer at Making Sense.
“That gives clients a public, verifiable reference for what we’ve delivered, which makes choosing who to work with less of a leap of faith.”
Before signing, a company can now compare firms on the Partner Hub by tier, certified team size, production deployments, and named references; questions a cautious buyer would otherwise have to dig into through reference calls.
Three months to launch, 6.6 cents an interaction
The firm’s published customer story is Doppler, an email and marketing automation company that serves businesses across Latin America and Spain. Making Sense designed and built ECO IA, an assistant that runs on Claude Sonnet 4.5 inside Doppler’s product.
ECO IA reads live account data, diagnoses how an account is performing and drafts campaigns. Users edit and approve each draft before it goes out, and the assistant launched within three months of kickoff. It is turned on by default for every new Doppler account, now available in more than 4,800 accounts, and 95% of users who engaged with it came back across multiple months, according to Making Sense.
For any company putting a model inside its product, the cost per interaction decides whether the feature helps or hurts margins. Making Sense used prompt caching, which stores the parts of a request that stay the same from call to call so Claude does not process them again each time.
The firm reports a 97% cache hit rate, and each ECO IA interaction costs $0.066.
Engineers also built a dashboard that tracks usage, costs, latency and errors in real time; Doppler can see what the assistant is spending and how it is performing as more accounts start using it.
From two months to a week
The second project shows Claude used inside the development process. At Viallion, Making Sense redesigned key parts of the engineering workflow around Claude Code, Anthropic’s coding tool.
According to the firm, initiatives that used to take two to three months to develop, validate and release now move in about a week, and ideas are validated four times faster.
At both Viallion and Doppler, Making Sense engineers worked inside the client’s team. At Doppler, the work included automated testing and changes to development workflows, so the in-house developers could keep testing, validating and extending the Claude integration on their own. Viallion’s team kept using and refining the new workflow after the enablement phase ended.
Cesar D’Onofrio, CEO and cofounder of Making Sense, said faster code generation has not removed the need for people who understand how a business runs, however.
“Claude has made building software dramatically faster. Knowing what a change will touch inside a company still takes people who understand that business: the system billing depends on, the data a regulator will ask about, the team that has to operate it the next morning,” he said.
“We’ve been doing that work for more than 20 years. We decide with our clients what is worth building, connect it to everything their operation runs on, keep it compliant and reliable, and answer for it once it’s live.”
Making Sense works mainly with U.S. companies, with a focus on agentic AI and workflow automation in regulated industries, including healthcare, fintech and logistics.
Featured image: courtesy of Making Sense
Disclosure: This article mentions a client of an Espacio portfolio company.
