Automation / AI Agent Operator
Application Task: AI Automation Specialist
We do not want to evaluate candidates based only on their CV or LinkedIn profile.
To apply for this role, choose one of the real operational problems below and build a working solution or prototype.
We are intentionally describing the problem, not telling you exactly what to build. A major part of this role is understanding an operational challenge, designing the right solution, and implementing it with minimal guidance.
Choose one problem
Option 1: Hiring Pipeline Automation
Clipster is hiring across several roles, but finding enough relevant candidates is time-consuming.
We currently use:
- Gem as our ATS and sourcing platform
- LinkedIn and other public sources for candidate discovery
- Slack for internal communication
- Job descriptions and evaluation criteria for each role
Design and build a system that helps us consistently bring more relevant candidates into our hiring pipelines.
Your solution may involve sourcing, candidate evaluation, data enrichment, workflow automation, notifications, or another approach you believe would solve the problem.
We are not looking for a simple workflow that only moves data from one tool to another. We want to see how you define the problem and design the overall system.
Option 2: Payment and Invoice Reconciliation
Clipster receives payments through several channels, including:
- Mercury
- PayPal
- Stripe
- Cryptocurrency
Invoices and financial records are managed through QuickBooks and internal documents.
Currently, matching incoming payments with the correct invoices, clients, and transactions requires manual work.
Design and build a system that can:
- Collect payment information from different sources
- Match payments with the correct invoices
- Identify unclear or unmatched transactions
- Provide a clear overview for the finance team
You may use test data or mocked integrations if direct access to a platform is unavailable.
Option 3: Internal Company Intelligence
Important information at Clipster is distributed across:
- Slack
- Meetings and meeting transcripts
- Emails
- Project management tools
- Internal documents
This makes it difficult for team members to understand what happened across the company during the week.
Design and build a system that identifies meaningful updates and produces a useful company brief.
For example, the system might capture:
- Product launches
- Engineering progress
- Important decisions
- Hiring updates
- Client developments
- Risks and blockers
- Changes in priorities
The goal is not to summarize everything. The goal is to identify what is actually important and present it to the right people.
Option 4: AI Pull Request Review
Our engineering team works on a fast-growing product and wants to improve the speed and quality of code reviews.
Design and build an AI-assisted pull request review system.
The system should go beyond generic code comments. Think about how it could use:
- Repository context
- Internal engineering documentation
- Product requirements
- Existing architecture
- Previous code patterns
- Team-specific standards
You may use a public repository, your own repository, or a mocked project environment.
What to submit
Please provide:
- A working prototype or demonstration of your solution.
- A short video walkthrough explaining how it works and why you chose this approach.
- A link to the project, such as a GitHub repository, hosted demo, workflow export, or shared document.
Please also briefly mention:
- Which tools and AI models you used
- What you would improve with access to Clipster’s real systems and data
Using AI is expected. We are evaluating how effectively you use it, not whether you complete everything manually.
What we evaluate
We will primarily assess:
- How well you understand and structure the problem
- Whether your solution addresses the underlying business need
- Ownership and ability to work without detailed instructions
- Quality of implementation
- Practicality and scalability
- Ability to connect multiple tools and data sources
- Error handling and edge cases
- Clarity of your explanation
- Speed and resourcefulness
A polished production system is not required. A thoughtful and functional prototype is more valuable than a large but superficial project.