Skip to main content

Marketing & Adtech · AI Outbound

AutoEngage AI

A custom AI-driven tool to automate and enhance client's LinkedIn engagement strategy and automate their networking

Linkedin AI Automation tool
Focus
A.I, INTERNAL-TOOLS, MVP
Delivered
17 Oct 2024
Location
Zurich, Switzerland

Outcome

Comments within 30 minutes of a relevant LinkedIn post, in the client's own voice.

Project background

Our client, Milasevic, approached Robust Devs with a forward-thinking concept: to develop a custom, AI-powered tool specifically designed to streamline and enhance their professional engagement activities on LinkedIn. Recognizing the critical role of consistent interaction for building networks and maintaining visibility on the platform, the client sought a solution to automate the process of commenting on relevant posts.

The objective was to create a Minimum Viable Product (MVP) that could intelligently generate human-like comments, target specific content using keywords, and engage with key profiles, thereby saving valuable time while maintaining an authentic and active online presence.

The challenge

Before partnering with Robust Devs, the client faced several obstacles in managing their LinkedIn engagement effectively:

Time Constraints: Manually searching for relevant posts, crafting thoughtful comments, and engaging consistently required a significant time investment, detracting from other core business activities.

Inconsistent Activity: Maintaining a regular and impactful presence on LinkedIn proved difficult, leading to missed opportunities for networking and visibility, especially during busy periods or time offline.

Scalability Issues: Scaling engagement efforts manually to cover desired keywords and key profiles was challenging, limiting the potential reach and impact of their networking strategy.

Maintaining Authenticity: Ensuring that each comment was relevant, personalized, and maintained a consistent professional tone demanded considerable effort for every interaction.

Engagement Timing: Responding quickly to new, relevant posts is crucial for visibility on LinkedIn, something difficult to achieve consistently through manual monitoring.

Compliance Concerns: The client recognized the need for any automated solution to operate within LinkedIn's terms of service to avoid potential account restrictions.

How we solved it

Our approach combined strategic planning with targeted technical implementation:

AI-Powered Comment Generation: We integrated OpenAI's advanced GPT-4 models (GPT-4o/GPT-4 mini) to generate high-quality, contextually relevant, and human-like comments. An initial onboarding process captured the client's industry, interests, and goals to create a personalized base prompt for the AI, ensuring comments aligned with their professional persona.

Targeted Engagement Automation: The application was engineered to automatically identify and engage with relevant LinkedIn posts.

Timely and Strategic Interaction: We implemented a scheduling mechanism ensuring comments were posted within 30 minutes of a relevant post's publication, maximizing visibility.

Intelligent Features: The system incorporated AI-driven language detection to ensure comments matched the language of the original post, maintaining authenticity across diverse interactions.

Iterative MVP Development: Following a structured, milestone-based approach (Design, Development, Testing & Deployment), we delivered the core functionality within the planned 1-1.5 month timeframe, allowing for focused development and refinement based on clear objectives.

A one-hour timeline from the moment a relevant LinkedIn post is published. Within a couple of minutes the keyword and followed-profile watch has seen it. By minute nine the post has been read and its language detected. By minute eighteen a comment has been drafted in the client’s own professional voice from their onboarding profile. The comment is posted before the thirty-minute mark, which is the window where visibility is highest. Manual replies typically land closer to the hour, by which point the post has stopped surfacing.

Technologies

What we built it with

Front-end
  • JavaScript
  • ReactJS
  • HTML5
  • CSS3
Back-end
  • PHP
  • Laravel
  • Python
DevOps
  • Gitlab CI
  • Digital Ocean
Database
  • MySQL
Integrations
  • OpenAI

What shipped

AutoEngage AI watches LinkedIn for posts matching the keywords and profiles the client chose, drafts a comment with GPT-4o from a base prompt built during onboarding, and schedules it to publish within thirty minutes of the original post. Language detection matches the comment to the language of the post it answers.

The MVP was delivered inside the planned one to one and a half month window, in the design, development, testing and deployment sequence agreed at the start. The client keeps a consistent presence on LinkedIn without spending the day on it, and every comment goes out in their own voice.

How we built it

Captured the client's professional voice once at onboarding, then let a scheduler beat the feed to the comment. A comment only counts on LinkedIn if it is early and sounds like the person. Onboarding records the client's industry, interests and goals into a base prompt, and GPT-4 class models write each comment against it, with language detection so the reply matches the language of the post. The application finds relevant posts on its own and the scheduling mechanism posts within 30 minutes of publication, which is the window where a comment is still seen. The MVP shipped inside the planned one to one and a half month window.

In the client's words

“Precisely the intelligent automation I was looking for! Saving me significant time managing my LinkedIn engagement and generating quality comments”

NDA - Signed

AutoEngage AI

More case studies

Have a project like this in mind?

Tell us what you're building. We'll show you how we'd approach it and what it would take to ship.

Schedule a meetingSee all case studies