How generative AI is reshaping the way people work.
Spawnex is a non-profit, independent research organisation studying the effect of generative AI assistance on employee productivity, work quality, and wellbeing across companies and sectors worldwide. We publish our findings openly and do not offer consulting or advisory services.
We study people, not just models.
Spawnex investigates how generative AI tools change day-to-day work — the tasks people delegate, the time they recover, the quality of what they produce, and the effect on burnout and engagement.
- Founded2023, Dakar
- StatusNon-profit research organisation
- FocusGenerative AI in the workplace
- MethodField studies & longitudinal surveys
- Sectors studiedRetail, banking, construction, technology, heavy industry
- DirectorSergey Voropaev
Based in Dakar, Spawnex runs applied studies inside partner companies rather than relying on lab simulations alone. We track how employees actually use AI assistants over weeks and months — not just how they perform in a single test session.
Our work sits at the intersection of organisational psychology and applied AI: we're as interested in employee engagement and wellbeing as we are in raw output metrics. Recent projects have combined generative AI adoption with reduced-hours work models to see whether the two reinforce one another.
Findings from our studies have been published in peer-reviewed technical journals and are made available in full below.
Three questions guide our current work.
Does AI assistance actually save time, or just move it?
We measure task completion speed, throughput, and error rates in AI-assisted versus unassisted workflows, using controlled and field comparisons within the same organisation.
What happens to accuracy when speed increases?
Quality is scored through structured expert review, 360-degree feedback, and error-frequency analysis — not self-reported impressions of AI usefulness.
Can AI support shorter, healthier working weeks?
We study engagement (via the UWES-9 scale), burnout, and absenteeism alongside reduced-hours pilots to see whether AI-supported schedules hold up over time.
What the data has shown so far.
Figures drawn from Spawnex field studies across companies in Eastern Europe, Switzerland, and Hong Kong, and from our synthesis of comparable international research.
Three years inside major energy retailers.
Since 2023, Spawnex has conducted a continuous research study inside large energy retail groups, following the adoption of generative AI from an initial pilot through to organisation-wide use across customer operations, forecasting, and field service.
Customer contact-centre AI assistant
The partner organisation's own generative-AI assistant, used alongside human agents for billing, tariff, and outage-related enquiries, was studied continuously over three years rather than in a single pilot window.
Demand forecasting & field-service scheduling
Working alongside the operations and analytics teams, Spawnex studied the organisation's use of generative AI in demand-forecasting workflows and dispatch scheduling for field technicians, with results validated against three consecutive annual cycles.
Papers from the lab.
Full texts are hosted on our shared drive. Both studies were led by Spawnex director Sergey Voropaev and published in Universum: Technical Sciences.
Integration of generative AI and reduced working hours: a synergistic approach to enhancing workplace productivity and employee well-being
A synthesis of multinational data, including the J-HOPE study, the Microsoft Japan trial, and the UK four-day-week pilot, examining how pairing AI assistance with a shorter working week affects productivity, quality, and burnout.
Symbiotic enhancement of employee productivity using generative AI in the business sector
An 86-participant field study across Eastern European, Swiss, and Hong Kong companies comparing two modes of AI collaboration — a collaborative agent and a continuous agent — against a control group on speed, completion rate, and quality.
How we run a study.
- RecruitPartner companies nominate employees; stratified randomisation balances experience, industry, and baseline productivity.
- AssignParticipants are split into AI-assisted and control groups, with two AI interaction modes tested where relevant.
- MeasureTask speed, completion rate, and quality are scored through timed tasks, expert review, and validated engagement scales.
- ValidateBootstrapped confidence intervals, regression, and structural equation modelling check that results hold up.
Led from Dakar.
Sergey Voropaev is an engineer and the founder of Spawnex, where he leads the lab's studies on generative AI adoption in the workplace. His research combines quantitative fieldwork with organisations across Europe and Asia to measure the real-world effect of AI assistance on productivity, quality, and employee wellbeing.
He has published his findings in Universum: Technical Sciences and continues to run applied trials with partner companies from Spawnex's base in Dakar, Senegal.
Take part in a study.
We collaborate with organisations that want to take part in a structured, independent study of AI adoption inside their own teams. There is no fee — participation is a research partnership, not a paid engagement.
Office
Spawnex Research Lab
Plateau District, Dakar, Senegal
cso@spawnex.com
Research partners
We welcome new research partners in retail, banking, construction, technology, and heavy industry who are interested in taking part in a study of AI-assisted or reduced-hours workflows.