Section 4.2: AI Matchmaking Suggestions

ABRAM features an advanced, AI-powered project matchmaking system designed to suggest the optimal crew members for your projects. Instead of searching and reviewing profiles manually, the matchmaking engine automatically analyzes project requirements and compares them against your team's real-time availability, skills, portfolio experience, budget, and working preferences.

1. The Matching Workflow

The matchmaking engine works on a per-role basis to compile optimal crew suggestions:

1. Splitting into Role Slots

A project’s work package is broken down into individual Role Slots (e.g., Cinematographer, Gaffer, Key Grip). Each role slot has a defined start date, end date, and required skills.

2. The Hours Allocation Sequence

Before matching, the engine determines the required hours for each role slot using a simple sequence of sources:

• Explicit Allocations: If you have already specified hours for a role within the deliverables (e.g., Editor: 15 hours, Colorist: 5 hours), the platform uses these values. This manual input is completely free.
• AI Estimation: If you haven't entered manual hours, you can use the AI Assistant to estimate the effort based on the project scope. To optimize credit usage, the system saves these estimations, so they are only calculated once.
• Even Split: If manual inputs are not specified and you do not run AI estimation, the platform splits the project's total estimated hours evenly among all active roles (e.g., 30 total hours split among 3 roles results in 10 hours each).

3. Effort Hours to Weekly Capacity Conversion

Once the total effort hours are determined, they are converted into a weekly planned capacity hold for scheduling:

• Short Projects (1 week or less): The weekly capacity hold is equal to the total effort hours.
• Long Projects (more than 1 week): The weekly capacity hold divides the total hours by the number of weeks, rounded to the nearest whole hour.

This value is stored as the proposed hours per week on the crew invitation.

4. Calendar and Booking Capacity Holds

Upon invitation acceptance:

• The system automatically creates a calendar booking marked as a Project Work Capacity Hold.
• Visual Layout: This booking appears as a neat, all-day banner at the top of the freelancer's calendar rather than blocking off specific hourly time slots.
• Freelancer Autonomy: This ensures scheduling availability checks remain accurate while giving freelancers complete autonomy to decide exactly when during the week they will perform the work. Freelancers log their actual hours worked on their weekly Time Card.

2. Crew Suitability Evaluation (0–100%)

Candidates are ranked using a comprehensive matchmaking algorithm that calculates a suitability percentage based on four major factors:

1. Technical Skill & Expertise Fit

• Skill Matching: The AI compares the required project skills against the skills listed on the candidate's profile. It uses synonym mapping (for example, if a project requires "Premiere Pro" and the freelancer listed "Adobe Premiere", the AI automatically recognizes this match).
• Software Proficiency: Checks familiarity with required production software tools.
• Role Alignment: Confirms whether the freelancer's primary declared roles match the slot.
• Equipment Matching: Checks if the freelancer owns or operates specific technical equipment required for the shoot.
• Specialization Fit: Evaluates whether the freelancer holds verified specializations related to the project type (e.g., video editing, motion design).
• Expertise Level: Considers the freelancer's average expertise level in their verified skills.

2. Location & Work Mode Fit

• On-Site Roles: For physical, on-location roles (like Gaffer or Cinematographer), the algorithm checks travel feasibility. It prioritizes local crew to minimize travel overhead, mileage costs, and accommodation logistics.
• Remote-Friendly Roles: For digital or post-production roles (like Editor or Designer), physical location is ignored. The engine instead evaluates timezone overlap to ensure smooth communication during collaborative windows.

3. Real-Time Availability & Capacity

• Schedule Analysis: Rather than relying on a static availability flag on a profile, the algorithm queries all active bookings in the candidate's schedule for the project's exact date window.
• Remaining Hours: The system subtracts current project commitments from the freelancer's maximum weekly capacity. Freelancers with sufficient unbooked time to cover the role's weekly requirements are ranked higher, while overbookings lower suitability.

4. Budget Alignment

• Rate Check: Compares the freelancer's declared hourly or daily rate against the target budget allocated for that specific role slot.
• Budget Fit: Freelancers whose rates fall within or below the budget range are prioritized, while rates exceeding the target budget will lower the candidate's suitability ranking.

3. Reviewing Suggestions & Concerns

To view AI matchmaking suggestions for a project:

1. Navigate to Projects and open the specific project dashboard.
2. Click Find Matches in the upper right. The engine will evaluate candidates for each defined role slot.
3. Review the Role Slot Matching Table which displays suggested candidates sorted by match score.

Match Reasoning & Concerns

Under each candidate's score, the interface lists:

• Match Reasoning: A quick summary of their strengths (e.g., "Strong fit with excellent technical skill match and high availability").
• Concerns / Red Flags: Potential risks, such as budget mismatches (hourly rate exceeds target budget) or timeline overlaps (conflicts with existing booked projects).

Once you've selected the optimal candidates, you can check their names and click Invite All Selected to dispatch invitations immediately.

4. Credit Consumption & Caching for Matchmaking

Running the AI matchmaking engine to analyze suitability, calculate scores, and generate match reasoning consumes platform credits from your workspace billing ledger.

Matchmaking Credit Rules

• Free Operations: Browsing the freelancer list, searching your internal registry manually, or viewing freelancer profiles does not consume credits.
• Credit-Gated Operations: Running the AI matchmaking suggestions (which calculates suitability scores and generates match reasoning or concerns) always consumes credits. This applies even if you manually entered the role hours or used the even split fallback instead of AI hours estimation.
• AI Processing: The system uses specialized analysis engines to evaluate candidates, identify compatibility concerns, and write detailed suitability rationales. Credit consumption is based on the volume of data analyzed per query.
• Onboarding Free-Tier: If you are a new organization founder completing your first-time onboarding setup, credit consumption is waived for your initial matchmaking trials.

Caching Safeguards

To protect your workspace budget from redundant credit charges:

• Saved Role Estimates: Once the AI estimates hours for a work package, the results are saved directly to the project's deliverables. Reloading the dashboard or reviewing the saved estimates does not consume additional credits.
• Match Reasoning Cache: The detailed match reasonings and concerns are cached for your session. Opening a candidate's profile preview or reloading the matching grid does not trigger a new credit charge. You only consume credits when you explicitly trigger a new search or re-evaluate matchmaking after changing the project's dates, roles, or deliverables.