the problem
Lebanese tech talent is world-class. but the job market infrastructure hasn't kept up. local platforms are outdated. international platforms don't understand the Lebanese context: the salary expectations, the timezone advantages, the diaspora network, the specific skill sets coming out of AUB and LAU.
talentap's founder saw a clear gap: a platform purpose-built for this talent pool and the employers who want access to it.
the problem for us was design-level: how do you build a two-sided marketplace that serves both candidates and employers well, without making either side feel like they're on a generic job board?
two users, one product
the sprint discovery week revealed that the two user groups had almost nothing in common in terms of motivation.
candidates wanted:
- to be found, not to search
- to know the salary range before applying
- to understand the company culture before wasting time on a call
- to feel like a professional, not a commodity
employers wanted:
- pre-vetted candidates only
- fast response times
- confidence that the candidate would actually show up to the interview
- clarity on availability and remote vs. relocation preference
the insight that shaped the whole product: both sides wanted to feel like the other side was serious. the design had to signal quality and mutual respect from the first screen.
the matching system
we designed the matching flow around a mutual-interest model. candidates don't apply to jobs. employers don't browse resumes. both sides express interest, and a match is made only when interest is mutual.
this required:
- a candidate profile that reads like a portfolio, not a CV
- a job listing format that shows culture and team, not just requirements
- a match notification UI that feels like an introduction, not a ping
- a lightweight messaging layer for the first conversation
the design language
talentap needed to feel premium without feeling exclusive. we built a design system around:
- a strong typographic hierarchy using Urbanist (weight contrast between job titles and metadata)
- a card system that surfaces the most trust-building information first
- a blue and dark navy palette that reads as professional without being corporate
2
user types
32
screens designed
30
days sprint
outcome
the prototype was validated with 20 candidates and 8 employers during the sprint. the mutual-match model was the single most positively received design decision. candidates reported feeling respected. employers reported better response rates compared to traditional outreach.
the platform launched in Q1 2024 and placed its first cohort of candidates within 6 weeks of going live.