MyMonday
Building an AI-powered skills graph and talent matching platform for fairer, more transparent hiring.
Challenge
Hiring, contracting and project staffing are often treated as keyword search problems: companies describe a role, candidates describe their experience, and platforms attempt to connect both sides through titles, filters and exact-match terminology. But real talent matching is more complex. People use different language for similar capabilities, skills are connected to adjacent skills, and experience only becomes meaningful when understood in context.
MyMonday needed to move beyond traditional marketplace logic and create a platform that could understand relationships between people, skills, roles, industries, companies and project requirements. The goal was not simply to help companies find candidates faster, but to create a fairer and more transparent foundation for matching people with opportunities based on actual capability.
Solution
FirstEnvision partnered with Firefly Information Management as a full product and technology partner for MyMonday, supporting the platform from discovery and UX/UI design through architecture, development, AI/ML matching logic, infrastructure, testing and deployment.
At the core of the solution is a Neo4j-based graph architecture with a custom graph layer. Instead of storing skills and profiles as flat records, the platform models them as connected entities: people, skills, related skills, positions, projects, companies, industries and experience. This allows MyMonday to calculate semantic fit rather than relying on exact keyword matches.
The platform also includes a custom AI/ML matching engine built without dependency on modern large language models. That made the technical challenge more demanding: semantic understanding had to come from structured domain modeling, relationship weighting, graph traversal and purpose-built algorithms rather than generic prompting. The result is a scalable talent intelligence platform capable of supporting external contracting, internal talent discovery and enterprise innersourcing workflows.
2016–2021
Main active delivery period
7 ms
Graph-based candidate-to-position match
1M × 1M
Positions vs candidates in benchmark scenario
Graph speed, measured
Matching throughput
Graph traversal against a relational baseline on the same matching workload.
≈143× faster
Neo4j graph traversal — custom graph layer
≈143 matches/s
Relational database — same benchmark
≈1 match/s
Internal benchmark: matching one million positions against one million candidates, complex candidate-to-position fit calculation on a single node.
How we built it
Product Discovery & Strategic Definition
FirstEnvision helped shape MyMonday as more than a talent marketplace. The early phase focused on understanding the business model, the hiring problem, contractor and company workflows, and the long-term product opportunity around skills-based matching. The strategic foundation positioned the platform around capability, transparency and semantic fit rather than keyword search or traditional CV filtering.
UX Architecture & Interface Design
The platform needed to make advanced matching technology feel practical and understandable. FirstEnvision designed workflows for contractor profiles, company accounts, onboarding, project creation, position definition, skill inventories, shortlists and match review. The UX challenge was to translate graph-based logic and AI/ML scoring into clear product interactions that users could trust in real hiring and staffing decisions.
mymonday.co
Capability-led brand and interface language, from the first screen on mymonday.co
Capability-first profile creation — skills and levels instead of a CV upload Graph Database Architecture
A graph-based architecture was selected because the core product problem depends on relationships. MyMonday needed to understand how skills relate to other skills, how people connect to experience, how positions combine requirements, and how companies define project needs. FirstEnvision implemented the foundation using Neo4j and a custom graph layer to model people, companies, projects, positions, roles, industries and skills as connected entities.
Semantic match pathAdjacent capability via graphContext relationships Custom Graph Layer & Semantic Data Modeling
FirstEnvision developed a custom graph layer to control how relationships are created, interpreted, weighted and used by the matching engine. This layer made it possible to represent partial matches, adjacent capabilities and transferable skills instead of relying only on exact term matching. The data model became the foundation for a more explainable and flexible talent intelligence system.
Application services
Marketplace workflows, matching API, onboarding, messaging, admin
Custom graph layer
Relationship creation, interpretation, weighting & explainability
Neo4j graph database
People, skills, positions, projects, companies as connected entities
AI/ML Matching Engine Without LLMs
One of the most important technical achievements was the custom AI/ML matching engine. Built before the current generative AI wave, the system did not rely on large language models to interpret text or produce recommendations. Instead, it used structured skill semantics, graph relationships, machine learning logic and custom weighting to calculate candidate-to-position fit. This created percentage-based matching that could support more transparent decision-making for companies and contractors.
Input
Contractor profile
Skills, proficiency, experience, availability, rate
Input
Position requirements
Required skills, seniority, project context, timeline
Matching engine — no LLMs
Graph traversalRelationship weightingStructured skill semanticsCustom ML scoringSemantic fit computed from the skills graph — including partial and adjacent matches.
96%Explainable
match scoreMarketplace & Enterprise Product Development
MyMonday was built as a complete digital platform, not only an algorithm. FirstEnvision delivered company and contractor workflows, onboarding flows, project and position creation, candidate matching, shortlists, messaging, admin functionality, payments and contracts, LinkedIn profile import and CV export. The platform also supported enterprise-oriented use cases such as Active Directory onboarding and internal talent discovery.
mymonday.co
Defining a position — the graph suggests adjacent skills and previews matches mymonday.co
Hiring pipeline from first match to signed offer Internal Talent Discovery & Innersourcing
Beyond external hiring, MyMonday supported internal talent discovery. This allowed organizations to identify employees who may fit future projects based on skills and experience already available inside the company. For larger businesses, this created a path toward better workforce planning, reduced dependency on manual recommendations and more effective use of existing talent.
Infrastructure, Testing & Launch
FirstEnvision provided infrastructure, deployment, testing, performance optimization and long-term technical ownership. The platform was publicly launched and validated through at least one B2B pilot client. Internal benchmark scenarios showed the graph-based matching approach calculating complex candidate-to-position matches in approximately 7 milliseconds, compared with approximately 1 second for a relational database approach in the same context.
The product
Product visuals re-created from the original 2016–2021 platform for presentation clarity.
Outcome
MyMonday demonstrates FirstEnvision’s ability to build technically ambitious digital products where product strategy, UX design, architecture, AI/ML and infrastructure must work together. The platform created a future-ready foundation for semantic talent matching years before skills graphs, knowledge graphs and AI-assisted talent intelligence became mainstream industry language.
By combining Neo4j, a custom graph layer and purpose-built matching algorithms, FirstEnvision helped turn a complex hiring challenge into a scalable product experience. The result is a publicly launched platform that supports external contractors, enterprise workflows and internal talent discovery through a fairer, more skills-based approach.
Nineteen years of craft — code and campaigns — at an honest price for honest work. Bring us the problem.