What do you use today instead of Dejo (https://dejo.run/) (a competitor, a spreadsheet, doing it by hand)? Compared with that, what would make Dejo better or worse for you, and what would keep you on your current option? Also say the maximum you would pay per month in USD (0 if nothing).

Flora Goyetteneutral

Right now, I'm mostly using a shared spreadsheet for competitor monitoring – competitor product pages, promo landing pages, stock notices, that sort of thing. I also use a few no-code automations and sometimes a one-off script from a dev friend to repeat browser tasks. The idea of Dejo learning a site once and then replaying it with zero model calls sounds brilliant, especially for cutting down on repetitive actions and tokens. That's a massive plus. However, it's open-source and self-hosted which means I'd need to figure out the setup and ongoing maintenance, which is a bit of a faff. I'd be concerned about how robust it is for dynamic sites or if a website changes drastically – does it then require a full relearn or just a minor fix? The benchmark showing significant savings on repeated tasks is promising, but the initial 'learning' cost is still a factor. For now, I'd say my current manual/scripting methods are more predictable, even if labour-intensive. I'd be willing to try it out if there was a very low-cost entry point, perhaps $10 per month to start, to see how it integrates with our existing workflow.

Simulated interest: 40/100

Lloyd Gislasonneutral

Right now, I'm piecing together custom Python scripts with Claude Code and Cursor, basically babysitting them to scrape data from client portals. I keep repeatable steps in notes and spreadsheets. The main advantage Dejo offers is the 'learn once, replay' concept, which sounds like it would significantly cut down on prompt engineering and debugging when websites inevitably change. That's a huge pain point for me. What I'm unclear on is how robust the 'repair what moved' functionality is when the DOM structure undergoes more significant shifts. I'd need to see that in action. Pricing isn't clearly laid out, but if it solves the core problem, I'd consider paying up to $20 per month.

Simulated interest: 60/100

Meena Gandhineutral

Right now, I spend most of my day manually reconciling orders across different supplier portals, emails, and spreadsheets. It's quite tedious. This 'Dejo' sounds interesting because it claims to learn websites and then replay actions with zero model calls. If it truly automates the repetitive browser confirmations and data copying I do, that would be a significant improvement over my current manual process. However, I'm concerned about how well it handles changes on those supplier websites, as that's a constant issue for me. I would also need to understand how it integrates with our existing systems. I'm willing to pay a maximum of $20 per month for this, provided it significantly reduces my manual effort without introducing new complexities.

Simulated interest: 60/100

Yusuf Bertenbreiterneutral

Currently, we rely on Playwright for our end-to-end testing, storing reusable scripts in GitHub and performing manual checks. The promise of Dejo remembering a website after the initial learning phase is intriguing, especially the claim of zero model calls on subsequent runs, which could significantly reduce costs compared to our current approach. However, I'm concerned about how it handles dynamic elements and potential flakiness, as that's a constant challenge for us. The benchmark shows potential savings, but I'd need to see how it performs on our specific complex checkout flows. Since it's open-source and free, my maximum would be $0 per month.

Melinda Klockoneutral

Right now, I manually copy candidate details between LinkedIn, Greenhouse, Calendly, and a spreadsheet. It's incredibly time-consuming and prone to errors. The idea of Dejo learning a website once and then replaying actions with zero model calls sounds appealing, especially if it means fewer manual steps. However, I'm not entirely clear on how it integrates with my existing workflow or if it can handle the complexity of switching between different sites like LinkedIn Recruiter and Greenhouse seamlessly. I'd need to understand its limitations and how much setup is involved. I wouldn't pay anything right now, as I'm still experimenting with basic Zapier workflows and need to see the value first.

Simulated interest: 30/100

Agilbert Meyerneutral

Currently, my teammate and I manage client website tasks using a combination of Notion for tracking and Google Sheets for analysis, augmented by Zapier automations. We also do a lot of manual browser work for tasks that can't be automated. Dejo's promise of an AI agent that remembers websites and reduces model calls sounds intriguing, especially the idea of learning a site once and then replaying it with a single tool call. This could potentially streamline our repetitive website verification and updates. However, I'm concerned about how it handles complex, dynamic client sites that change frequently, and what the actual setup and learning process entails for a new site. The open-source nature is appealing, but I'd need to understand the support and maintenance involved. Given our lean operation and focus on reliability, a significant migration effort or fragile automation would be a non-starter.

Simulated interest: 40/100

Melissa Tillmanneutral

Well, I currently use spreadsheets and my own two hands for most of it, which can be a bit of a faff, especially when updating listings across different marketplaces. The idea of an AI agent remembering websites sounds intriguing, saving me time on repetitive tasks. However, I'm a bit hesitant about needing to fix things when websites change. I'd need to see how that actually works in practice. It also mentions using it with Claude Code, Codex and Cursor – I'm not entirely sure what those are or if they'd work with what I have. The biggest hurdle is the cost; I'm quite frugal. I'd be willing to pay up to $10 per month, but only if it genuinely saves me significant time and hassle. The pricing itself isn't clearly laid out here, just that it's a monthly subscription in USD.

Simulated interest: 30/100

Rex Gibsonneutral

Currently, I rely on a mix of manual browser steps, shared macros, and a lightweight RPA tool to manage repetitive account admin tasks across different internal and external platforms. My current process is functional but time-consuming, and I'm always looking for ways to improve efficiency and reduce manual entry errors. The concept of Dejo remembering websites and reducing subsequent model calls is intriguing, especially the idea of a single tool call after the initial 'learning' phase. This could potentially streamline my workflow significantly. However, I would need to understand how it handles dynamic websites that change frequently and if the 'repair' process is truly seamless or adds new complexities. My primary concern is reliability and ease of use for someone in my role, not necessarily cutting-edge AI. Given my focus on pragmatism and cost-effectiveness, I would be willing to pay up to $20 per month, provided it genuinely saves me time and reduces errors without a steep learning curve or significant maintenance overhead.

Simulated interest: 60/100

Garuda Guptaneutral

Currently, I rely on a combination of Playwright scripts and custom Python code, sometimes augmented with Claude Code for specific AI tasks, to automate browser workflows for my clients. Doing things manually is obviously out for any repeatable process. Dejo's core promise of remembering websites to reduce model calls is intriguing, especially for repetitive tasks. The significant reduction in tokens and tool calls after the initial run, as shown in your benchmark, could be a major cost saver. However, my current setup is highly flexible, and I can integrate various open-source components as needed. The ability of Dejo to repair broken steps dynamically and maintain signed-in sessions is appealing. What would make Dejo worse is if it proved less adaptable than raw Playwright for complex, dynamic client websites, or if the 'repair' mechanism was too rigid. The biggest hurdle for adoption would be any steep learning curve or integration complexity for my existing client workflows. I'm always looking for efficiency, but I also need robust solutions that are easy to maintain. I'd be willing to pay up to $20 per month, primarily to experiment with its capabilities on a few select client projects.

Simulated interest: 60/100

Lien Terlouwneutral

Currently, we rely on spreadsheets and manual browser-path checklists for our release checks. One team member, in particular, replays the same flows repeatedly for every launch. The concept of Dejo, where an AI agent learns a website and replays it with minimal calls, is intriguing. This could potentially alleviate the repetitive task burden my junior colleague faces. However, it's unclear how it integrates with our existing admin consoles or handles exceptions and discrepancies, which are a significant part of our workflow. The "sign in once" feature sounds useful for managing multiple accounts. I'd need to see a pilot or more detailed documentation on how it handles complex, multi-step processes and error recovery before considering it. The pricing basis is per month subscription, but the actual cost is not listed, making it impossible to estimate an acceptable price.

Simulated interest: 30/100