The hiring bottleneck is structural
A growing company can attract hundreds of applications for one role and still struggle to identify the right person. CVs arrive in different formats. Job criteria are interpreted inconsistently. Interviewers ask different questions and record uneven notes. Speed becomes the enemy of care.
Small and medium-sized businesses feel this pressure most sharply because they may not have specialist recruitment teams. A founder or manager reviews applications between other responsibilities, increasing delay and the risk that strong candidates are missed. Candidates experience the other side: silence, unclear expectations and interviews that do not consistently test the work.
Vera is being built as an AI-powered HR assistant for applicant screening and live interviews. The product is coming soon. Its intended role is to help teams define criteria, organise applications, structure interviews and preserve evidence for human review. The final hiring decision must remain accountable to people.
Hiring technology should make judgement more visible, not hide it behind a score.
A labour market changing at scale
Sub-Saharan Africa has a young and expanding workforce. The ILO estimated a 21.9% youth NEET rate in 2023 and noted that only a small minority of young adults attain work that qualifies as decent. In Eastern and Southern Africa alone, the World Bank reported in 2026 that about eight million young people enter the labour market each year, while fewer than one million secure waged jobs.
Employers are changing what they need at the same time. Estimates cited by the World Bank suggest more than 230 million jobs in Sub-Saharan Africa will require digital skills by 2030. In Nigeria, Côte d'Ivoire and Rwanda, 35% to 45% of jobs are projected to demand digital competencies. A hiring system must help employers see relevant skills without treating polished CV language as a perfect proxy for ability.
Digital recruitment can widen reach. IFC research notes that online recruiting tools can lower recruiting costs and help employers reach candidates in other countries. But wider reach also produces more applications, which makes structured screening more important.
What Vera is being designed to support
The first layer is role definition. Before AI evaluates anything, a hiring team should identify the outcomes, skills and minimum requirements that genuinely matter. Vague job descriptions create vague screening. Vera should help convert those requirements into criteria that interviewers can understand and candidates can be assessed against consistently.
The second layer is assisted screening. AI can extract relevant experience, organise application information and flag where evidence is missing. It should not silently reject people based on unexplained patterns. Hiring teams need to inspect recommendations, adjust criteria and recover candidates who may have taken a non-traditional route.
The third layer is the live interview. Vera's direction includes structured questions, note support and evidence capture during interviews. Structure improves comparability: every candidate for the same role should have a fair opportunity to respond to the capabilities that matter. AI can help an interviewer follow up on incomplete answers, but it should not infer emotion, honesty or personality from facial movement or voice.
The final layer is decision documentation. A shortlist should show why candidates advanced, what evidence supports the assessment and where uncertainty remains. This makes the process easier to audit and helps a team learn whether its criteria predicted successful performance after hiring.
Bias is a product requirement
Hiring data reflects past decisions, including past exclusion. Training an AI system on historical outcomes can reproduce those patterns at scale. Names, addresses, schools, career gaps and writing styles may act as proxies for characteristics that should not determine opportunity.
Responsible design therefore requires testing across gender, age, disability, language and other relevant groups; monitoring selection rates; providing human appeal; and limiting data to what is job-related. Models and criteria will change, so evaluation cannot be a one-time launch exercise.
Africa's linguistic and educational diversity makes local testing particularly important. A system built elsewhere may misread qualifications, career pathways or communication patterns. Africa-first does not mean claiming that local data is automatically fair. It means accepting responsibility for performance in the context where the product is used.
Efficiency should be measured carefully
Time-to-shortlist and administrative hours saved are useful measures, but they are incomplete. Vera should also be judged by consistency between interviewers, candidate completion rates, reasons for withdrawal, representation through each stage, quality of hire and the number of decisions changed after human review.
False confidence is a major risk. An AI-generated score may look more objective than a person's notes while resting on assumptions the company has never examined. Vera's interface should surface evidence and uncertainty, using scores only where their meaning is clear and validated.
A tool for better human decisions
Africa needs both more jobs and better systems for connecting people to them. Vera cannot solve job creation, education quality or every form of discrimination. It can make one consequential process more structured and transparent.
The ambition is practical: help a growing company handle volume without reducing candidates to keywords; help interviewers focus on evidence; help leaders understand why a decision was made; and give applicants a process that respects their time. Speed matters. Accountability matters more. Vera is being built to hold both.