Launching today
Construct Computer
Your AI coworker gets a computer. You get your day back.
92 followers
Your AI coworker gets a computer. You get your day back.
92 followers
AI workforce for solo founders and small teams. Any MCP or skill installs like an app, agents build the tools that don't exist yet, deployed and managed by the platform, shareable to the whole team, and any agent action that you like turns into a reusable workflow so agents spend less time thinking and more time doing.
Products used by Construct Computer
Explore the tech stack and tools that power Construct Computer. See what products Construct Computer uses for development, design, marketing, analytics, and more.
LLMs 1
LLMs 1
ComposioYour agent acts. We'll handle the rest.
5.0 (7 reviews)
Also considered:
PipedreamBuilding OAuth for a thousand apps is not a product, it's a tax. Composio handles the auth dance and gives our agents real tool calls into Gmail, Slack, Notion, and everything else, so users connect an app in one click and it just works. It also partly powers our agents web search and live browsers.
Engineering & Development 2
Engineering & Development 2
Kilo CodeThe open-source agentic engineering platform
5.0 (46 reviews)
We ship every day, so the coding agent had to be one we could actually trust with a large change. Kilo Code's orchestrator mode splits it into subtasks instead of one long context that drifts by the end, it runs whatever model you point it at with no markup on the tokens, and it's open source, so nothing about how it works is a mystery to us.
SentryApplication monitoring and error tracking software
4.9 (86 reviews)
A lot of users trust Construct with their operations, and when something breaks, we instantly receive a stack trace in our DMs and emails, Sentry gives us the unified trace across the Worker, the Durable Object and the users client app, which means we're debugging one thing instead of five, and that makes our lives a whole lot easier.
Ecommerce 1
Ecommerce 1
Dodo PaymentsPayments & Billing for AI, SaaS, and Digital Products.
4.8 (25 reviews)
Also considered:
StripeBilling was the one thing we didn't want to think about, and Dodo made it disappear. It works across every major country and currency we care about, and because they act as Merchant of Record, taxes and compliance stop being our problem entirely. For a small team shipping fast, that's a whole function we never had to think hard about.
Social & Community 1
Social & Community 1
Cloudflare WorkersYour site. Our network.
5.0 (64 reviews)
Also considered:
Google Cloud Platform
AWSWe wanted one platform where compute, storage, and AI all live in the same place, and Workers is the only one that delivers it. Durable Objects, R2, D1, Vectorize, and Workers AI, even Sandboxes and Browsers are bindings rather than separate vendors with separate bills and separate latency. Every agent we run executes on the edge network, close to the user, with no cold start we ever have to apologize for.
Data analysis tools 1
Data analysis tools 1
PostHogLiterally every piece of SaaS that a product engineer needs
5.0 (255 reviews)
Also considered:
Google AnalyticsConstruct is one screen with windows on it. PostHog gave us events we actually define, funnels through agent setup, and session replay so we can watch where someone hesitates before handing off their first job. Flags and experiments run off the same events. For a small team, it's the most leverage you can get from one install.
Productivity 1
Productivity 1
FigmaThe collaborative interface design tool
4.9 (1.5K reviews)
Construct is an entire desktop OS - windows, glass surfaces, a launchpad, a dock. That's a lot of surface to get wrong. Figma is where we worked all of it out before writing any of it, including every frame in this gallery. The MCP server means the design comes back out as code.
General 1
General 1
Context.devOne API to scrape, enrich, and understand the web.
4.9 (15 reviews)
jinaai
FirecrawlOur agents live or die on what they can read from the open web, and most scraping APIs hand back a wall of HTML that the model then has to pay tokens to untangle. Context.dev returns structured, model-ready data instead, including from sites that aggressively block bots, so the agent spends its context window reasoning rather than parsing navigation menus. One API covers fetching, extraction, and enrichment, which replaces what would otherwise be a scraper, a proxy layer, and a cleanup step we maintain ourselves.