TL;DR - Matching

  • Smart Matching: Combines traditional exact matching with AI to find related skills (>50% similarity)
  • Scoring:
    • Overall score: 0-100% (Perfect → Strong → Good → Medium → Moderate → Low → None)
    • Weights: Skills 30%, Availability 20%, Other criteria 50%
  • Key Features:
    • Finds both exact and related skill
    • Checks partial availability matches
    • Level matching includes adjacent levels (50% score)
    • Other criteria need exact matches

Coming soon: Custom weights, skill proficiency, project experience analysis

At least one skill or certification required; partial availability matching possible


📜 Details

The matching engine combines rule-based logic with AI-powered skill analysis to find the best candidates. It processes multiple criteria and provides detailed scoring for each aspect.

The use of AI helps us to widen the search scope, to find people with related skills

Example: You might be searching for people with “Agile” as a skill

In traditional rule-based systems, you would rely on the skill being explicitly mentioned by the person.

In our combined approach, we can find both, the exact, explicit skill as well as related skills, that are similar

Example: The person might not have listed “Agile” as a skill, but “Agile Methods”, which has a high similarity and is therefore counted in

Scoring System

Overall Match Score

  • Normalized scale of 0-100% and concordant clusters
    • 100% = Perfect Match
    • 99-80% = Strong Match
    • 79-60% = Good Match
    • 59-40% = Medium Match
    • 39-20% = Moderate Match
    • 19-1% = Low Match
    • 0% = None
  • Weighted combination of all provided criteria
  • At least one skill or certification required
  • Other criteria are optional

Criterion Weights

  • Skills: 30%
  • Availability: 20%
  • Other criteria combined: 50%
    • Certifications
    • Service Portfolio
    • Industry Experience
    • Seniority Level
    • Languages

In a future release, we will include manual weighting

Matching Criteria

Skills

  • Exact matching of skills
  • In addition: AI-powered similarity analysis (via so called “embeddings”) ^061bdf
    • Identifies semantically related skills with >50% similarity
    • Similarity (e.g. 70% is added into the skill score)

Searching for "Agile"

It finds the people that explicitly have “Agile” as the skill.

Besides that it also identifies “Agile Methodologies” as a related skill (e.g. 70% similar) and includes people with these skills as well.

Benefit:

  • The employee does not need to list all its skills but can focus
  • The user gets a wider search result

Availability

  • Evaluates three components (all components required if any provided)
    • Start Date
    • End Date
    • Weekly Capacity
  • Day-by-day availability check
  • Score based on percentage of matched days
  • This enables to also find people who might not have the full, but partial availability

Additional Criteria

  • Certifications: Exact match required
  • Service Portfolio: Exact match required
  • Industry Experience: Direct comparison
  • Seniority Level:
    • 100% for exact match
    • 50% for one level difference
  • Languages: Exact match required

Upcoming Features

  • Inclusion of all flowcase (f.k.a CV Partner) data, i.e. project experiences to find hidden skills
  • Skill proficiency levels
  • Configurable criterion weights
  • Enhanced matching algorithms