The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
Autonomous QA exploration with BugSnap AI: it explores a staging app on its own, re-confirms every candidate bug before reporting, and ships evidence, not guesses.
Quality Assurance TestingAI Agent DevelopmentQuality Assurance EngineeringNext.jsPlaywrightTypeScript
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Related posts
One of our clients needed 70 ad video variants in a week, so we built an AI agent to do it in seven hours.
The build itself was a project. It required:
- Assembling a Claude agent connected to Remotion
- Building a knowledge base of the client's brand: brand guide, movement language, hook creation guidelines, past deliverables, briefing structure
- Teaching the agent to read the brief the way our team would
- Composing hook variants adapted for each aspect ratio (1x1, 9x16, 16x9)
- Assembling endcards and outputting files named and organized for QA
The first batch still needed heavy manual editing. The second, the agent assembled alone.
The third came out fully stitched (hook, body, CTA, disclaimer) needing only a review pass before shipping.
Each iteration meant testing what the agent got wrong, adjusting the skills, refining the knowledge base, and rerunning until the output matched the standard our team would ship.
When we sent the batch, the client feedback was: "The revised videos look great. We'll start using them immediately."
To show what this workflow can do, the video in this post was produced by this exact system.
If your business needs to ship ads at this volume, let's talk 👉 apta.agency/contact
70 ad variants in seven hours is wild. remotion?
Heron AI is an agent that reviews architectural models and flags building code problems. The hardest part of the brief was making that legible in the first few seconds.
We built the hero as a small interaction instead of a statement. The visitor drags across a black and white building sketch and violations surface where they drag, annotated with real IBC clauses. It puts the person in the agent's position for a moment.
Everything after that section can then talk in normal product language, because the core idea has already landed without needing to be described.
Really smart way to make a technical AI capability tangible—the drag interaction lets the user discover the value instead of reading a claim. The architectural visual language supports that clarity beautifully.
Northstar CRM: Full-Stack Customer Relationship and Sales Management Platform
I designed and developed Northstar CRM to transform customer relationship management into a focused, visually intuitive, and connected workspace. Instead of overwhelming sales teams with dense tables and fragmented tools, I structured the platform around clear customer insights, actionable sales data, and streamlined daily workflows.
The visual identity combines a soft atmospheric background with glass-inspired surfaces, dimensional cards, and vibrant blue, teal, yellow, and charcoal accents. Carefully layered shadows, subtle gradients, and elevated components create a refined 3D appearance while maintaining clarity and professional usability. The entire interface follows a consistent 4:3 presentation ratio, making every screen suitable for portfolio displays and high-quality screenshots.
The overview dashboard gives users an immediate picture of business performance. Key metrics—including total revenue, pipeline value, completed tasks, priority deals, recent activities, and upcoming meetings—are presented through scannable cards, progress indicators, compact charts, and contextual status colours. This allows teams to identify important changes and required actions without navigating through multiple reports.
The customer management screen provides a searchable directory with company information, recent interactions, account value, relationship status, and assigned ownership. Selecting a customer dynamically updates the detailed profile panel, where users can review contact information, relationship health, next actions, and communication options such as messaging, calling, or scheduling a meeting.
The deal pipeline organizes opportunities into clear stages, including Discovery, Proposal, Negotiation, and Closing. Each opportunity card displays its customer, estimated value, expected closing date, owner, and current stage. The Kanban-style structure gives sales teams a visual understanding of pipeline movement while supporting faster prioritization and opportunity management.
The calendar and reporting screens connect operational planning with performance analysis. The calendar provides monthly scheduling, customer meetings, deadlines, focus time, and team commitments. The reporting workspace combines revenue trends, conversion rates, customer retention, pipeline sources, account health, and period-based performance comparisons through interactive charts and summary cards.
On the backend, I implemented a modular Node.js and Express REST API connected to a PostgreSQL database. The data model supports users, customers, companies, contacts, deals, pipeline stages, activities, tasks, meetings, notes, and reporting records. Structured relationships and indexed queries ensure that customer histories, pipeline summaries, and dashboard analytics remain fast and consistent as the dataset grows.
Authentication is handled through secure token-based sessions, password hashing, protected API routes, and role-based access control. Different permissions can be assigned to administrators, sales managers, and sales representatives, ensuring users only access or modify information relevant to their responsibilities. Input validation, centralized error handling, audit logging, and rate limiting strengthen reliability and application security.
Real-time updates allow changes to deals, tasks, customer activity, and meeting schedules to appear across connected screens without requiring a manual refresh. Background processes handle reminders, notifications, scheduled follow-ups, and reporting calculations, while caching improves the performance of frequently requested dashboard metrics.
Across the complete full-stack platform, I prioritized reusable frontend components, responsive layouts, secure API design, reliable data relationships, and consistent interaction patterns. The result is a modern CRM experience that feels premium and approachable while providing the technical foundation required for scalable customer management, sales execution, and business intelligence.
The backend model is the interesting part here. Once customer history, pipeline stages, tasks, and meetings share relationships, it’s easy for dashboard numbers to drift from the underlying records. Did you calculate those summaries on demand, or keep derived reporting data updated through the background jobs you mentioned?