The visible 13 months sit on experience, capital, and a full team
Mukund Jha previously worked at Google and helped build Dunzo. He started Emergent with his twin brother Madhav Jha, the company's CTO, in June 2025. Their prior engineering, operations, hiring, and fundraising experience is part of the result rather than background that can be removed from the story.
TechCrunch reported in July 2026 that Emergent raised a $130 million Series C at a $1.5 billion post-money valuation, bringing total funding to $230 million. This is not a bootstrapped weekend project, and a reader should not translate the founders' current speed into an ordinary solo-builder baseline.
Customers paid to replace operational work, not to watch a demo
Emergent targets entrepreneurs and small or medium-sized businesses that still run operations through spreadsheets, email, and messaging apps. Examples reported by TechCrunch include shipment tracking, factory systems, construction workflows, property management, custom CRMs, ERPs, inventory tools, and logistics software.
The commercial mechanism is concrete: customers describe a workflow, the product generates and deploys an application, and Emergent charges through subscriptions, usage-based pricing, deployment, and hosting. The user buys a working operational result. An AI label or impressive first screen is not enough.
$120 million ARR is a company claim, not audited profit
The July report says the company reached $120 million in annual run-rate revenue and more than 200,000 paying customers. Four months earlier, Emergent had reported more than $100 million ARR. These are statements supplied by the company through founder interviews and media coverage; the public sources do not provide a complete audited income statement.
ARR annualizes a current revenue pace. It is different from cash collected, net revenue after refunds, gross margin, profit, valuation, funding, and founder income. The number should remain a self-reported ARR figure, and it should not be presented as a typical outcome for AI application builders.
The smallest transferable product is one spreadsheet workflow
Choose one task that a real business repeats every week, such as reconciling inventory, assigning incoming leads, preparing quotes, or tracking a shipment exception. Write the input, output, user permission, failure path, and acceptance rule on one page before generating any code.
Ask one operator what the delay or manual error costs and offer a narrow paid pilot. A compliment, signup, or polished demo is weaker evidence than access to a redacted sample, a scheduled trial, or a small payment. If the operator will not commit, change the workflow or offer before adding models and features.
Separate the source company's tools from an APIToken candidate test
The public reports describe a natural-language application builder and agent-based workflow, but they do not disclose a complete, verified list of foundation models behind Emergent. No model available through APIToken should be described as Emergent's actual stack without direct evidence.
Within the models currently and lawfully offered at https://APIToken.Company, create a separate project key, inspect the public status page, and run the same redacted workflow sample under a small written budget. Record usage, failures, retries, manual corrections, and whether the output meets the acceptance rule. Model visibility is not workflow success, and one successful call is not production readiness.
Source and evidence boundary
The main sources are TechCrunch interviews published on July 15 and February 17, 2026, plus the Y Combinator company profile. They support the founder identities, launch window, product category, reported ARR and customer counts, funding, user workflows, and revenue mechanisms.
Emergent, Mukund Jha, and Madhav Jha have no demonstrated use of or partnership with APIToken. The $120 million ARR and 200,000 paying-customer figures remain company-reported and unaudited here. APIToken is mentioned only as a controlled way to isolate a project key, inspect live status and usage, and bound a small real test.