Fly Labs
Ask FlyBot

Looking for suggestions on how to approach automation

Thought I would ask for a slice of your wisdom! As things are moving so fast, I'm not up to speed with the latest/best methods for automation, and I've got a task I'd like to have a go at creating, but thought I'd ask which toolchains you might approach it with - ironically I don't really trust the tools to give me a good answer! **Background**: I'm an Embedded Software/System engineer - and have had good experience developing a massively complex AI driven tool using GitHub CoPilot with Claude 3.5-4.6, LangChain and and LangSmith with RAG etc. - So I understand the background principles, but this 'Code' option I'm sure is way to complex for the nut I want to crack this time. Also had a bit of a trial with LangFlow - but didn't find it that practical although didn't give it much time. **Pre-requisites**: * Ideally I'd like to start off for free, or utilising what I'm already paying for (GitHub Copilot Pro), or redivert my $10 a month to something more useful/better * Not going to support OpenAI any more given their recent stance, but happy to support Anthropic (have found Claude superior anyway) * Would prefer to self-host/run in codespaces/run locally - this is going to be a 'once a week' automation - so should be basically free * If I have to use a paid hosted tool - at least want to be able to set up my automation/toolchain and test it before having to pay/subscribe to anything * Ideally 'no/low code' but of course coding isn't really a hurdle these days **Goal**: As a start-up business, networking is really important for us, but there are so many possible local events that we could go to, it takes a few hours a week to run through all the various listing pages to find new events, decide if they're relevant, then sign up. I think it should be possible to automate this. **Requirements**: * I want to be able to input/provide a list of 'events' websites to the workflow * That list should be scraped (not sure how to handle pagination...) for 'events' * Use a

RedditToolProductivitySource
0Sign in to voteCopy link

FL score

68

out of 100

Verdict

BUILD

high confidence

Competition

22

competitors found, growing market, big tech present, funded players

Trend

8 community mentions

A self-hostable, Anthropic-powered tool for solopreneurs to automate local event discovery and filtering from specified websites, with a focus on reliability and cost-efficiency.

The pain

Startup founders spend 'hours a week' manually browsing multiple event listing sites, deciding relevance, and signing up for local networking opportunities. Existing general automation tools are either too expensive, complex to set up, or prone to breaking for this specific, dynamic task.

The gap

There is a gap for a specialized, reliable, low-cost (preferably self-hostable) automation tool that combines web scraping (with pagination handling), LLM-driven relevance filtering (Anthropic-preferred), and automated sign-up, specifically designed for event discovery for solopreneurs. General automation platforms are too generic and costly for this specific use case.

Build angle

Develop a self-hostable 'event agent' that takes a list of URLs, uses robust, adaptable web scraping (perhaps a modular approach for common site structures) to extract event details, then employs Claude 3.5/4 to filter events based on user-defined criteria, and finally provides options for automated sign-up or streamlined manual action. Focus on a clear MVP: reliable scraping and filtering to generate a curated list first.

Strengths

  • Addresses a clear, time-consuming pain point for startup founders.
  • Highly specific niche with strong user constraints (Anthropic, self-host, low-cost) that incumbents don't serve directly.
  • User's strong AI/LLM background provides a competitive advantage in building the intelligent filtering component.
  • Growing market for AI and agentic automation.
  • Potential for significant time savings for target users.

Risks

  • Building a truly robust and generic web scraping solution (especially handling diverse site structures, pagination, and anti-bot measures) is technically complex and prone to breaking.
  • Automating event sign-ups adds another layer of significant complexity and potential failure points.
  • The 'free/low-cost' requirement makes monetization challenging and limits resources for ongoing maintenance and improvement.
  • Strong incumbents in general automation (Zapier, n8n) mean that a generic 'better automation' won't cut it; the niche must be very strong.
  • Validation is needed to confirm a broad enough market segment of solopreneurs willing to pay for this specific solution, given the DIY alternatives and the self-hosting requirement.

Next steps

  1. **Refine the MVP:** Focus on reliable, multi-site scraping and AI-driven filtering as the core value. Defer automated sign-up to a later stage.
  2. **Deep Customer Interviews:** Validate with 10-20 startup founders/solopreneurs: how much time they *actually* spend, their current workarounds, their willingness to pay for a dedicated solution (even a low-cost one), and their comfort level with self-hosting/setup.
  3. **Test Scraping Feasibility:** Create a proof-of-concept scraper for 3-5 diverse, common event listing sites to assess the true technical difficulty and maintenance burden.
  4. **Explore Monetization:** Given the free/low-cost preference, investigate 'freemium' models, one-time purchase options, or small monthly subscription for hosting/premium features/maintenance if the self-hosted version gains traction.

Questions about this idea?

FlyBot reads the scoring and gives you a second opinion on “Looking for suggestions on how to approach automation”.

Open FlyBot