I spent roughly a decade building and running L&D programs at growing companies before I founded Learnforge. First at a fintech in its early-growth phase, where we went from about 80 employees to over 400 in two years. Then at a regional healthcare network, where I inherited a learning function that served around 1,400 staff across multiple locations with very different training needs. Both experiences taught me the same lesson in different registers, and that lesson is the reason Learnforge exists.
The problem that kept showing up
At both organizations, the pattern was identical. We would hire a cohort of people for similar roles. We would put them through onboarding — a program we had built with care, refined over time, and genuinely believed in. They would complete it. And then, four to six months later, maybe half of them would be performing at the level we expected. The other half were still finding their footing. Some of them never did, and we would part ways, and the whole expensive cycle would begin again.
The confusing part, early on, was that the program looked fine by every metric we had. Completion rates were high. Post-training satisfaction scores were decent. The content had been reviewed by subject matter experts. The modules had been updated recently. On paper, the program was working.
What I eventually understood — and it took longer than I would like to admit — was that we were measuring the wrong things and asking the wrong question. We were asking "did people complete the training?" when we should have been asking "did the training close the gaps that were actually blocking their performance?" Those are completely different questions, and for most of the hires who struggled, the answer to the second question was no.
The uncomfortable specifics
Let me give you a concrete example from the fintech period. We were hiring customer operations specialists — a role that required a specific combination of financial product knowledge, regulatory awareness, and customer problem-solving skills. Our onboarding program for this role was 32 modules, built over two years, and covered all of those areas.
What we discovered, after running a basic skills diagnostic on a struggling cohort, was that the reason most of them were behind at 90 days was a specific gap in one area: understanding the regulatory context for the product category they handled. Not the product itself. Not the customer communication. A specific knowledge domain that the program technically covered — but covered in a way that was too abstract for people who came from non-financial backgrounds, which was increasingly common as our hiring pool expanded.
We could have fixed that specific gap in a focused 3-module deep-dive for people with non-financial backgrounds. Instead, we had been routing everyone through 32 modules, most of which were not relevant to that gap at all. And the 3 modules that were relevant were buried in a sequence that learners reached weeks into the program, by which time they were already behind.
The waste was staggering once I saw it clearly. Thirty-two modules of required content. Three modules that actually moved the needle for the population that was struggling. And no mechanism to identify which three, for whom, before wasting 29 modules of everyone's time.
Why the catalog is a symptom, not the cause
The catalog is not the villain in this story. L&D teams build catalogs because they are trying to be thorough in the face of uncertainty. If you don't know which specific skills your new hires will be missing, you cover everything. That is a rational response to an information problem.
The information problem is: most L&D teams do not have a mechanism to know, before they assign content, where the actual gaps are for a specific person in a specific role against a specific team baseline. Without that information, comprehensive coverage is the only defensible choice. You can't take risk on a gap you can't see.
The solution, therefore, is not to criticize the catalog — it is to build the information infrastructure that makes targeted content assignment possible. That means a skill taxonomy that is precise enough to be assessable. A velocity baseline derived from the actual performance distribution of your existing team, not a job description. A day-one assessment that produces a real skill profile, not a self-report survey.
With those three things in place, the catalog becomes optional enrichment. The assigned path becomes a targeted prescription.
Why I built this instead of consulting on it
By the time I left the healthcare network role, I had a fairly clear picture of what needed to exist. The pieces were not individually novel — adaptive assessment had been around for decades, skill taxonomies were well-understood, gap analysis as a concept had been in the L&D vocabulary forever. What did not exist, in a form accessible to an L&D team at a 500 to 5,000-person company, was a system that tied them together automatically and surfaced the output in a form that was actionable for both the hire and the manager.
Every vendor I evaluated at the time solved part of the problem. LMS platforms handled content delivery but had no gap analysis layer. Assessment tools gave you a skill profile but no path assignment logic. Competency management platforms handled the taxonomy but were priced and designed for large enterprises with six-figure implementation budgets and dedicated HRIS teams.
The mid-market gap was obvious. Growing companies with 500 to 5,000 employees — the exact cohort where L&D investment pays the highest return and where the bandwidth to build custom systems is the most constrained — were being served by either enterprise tools that required too much setup or consumer tools that lacked the depth. Nothing was built for the L&D team of two trying to support 200 new hires a year with precision and without burning out.
That is the gap Learnforge exists to close. Not the catalog — your people still need good content. The gap between your hire's day-one skill profile and your team's velocity baseline, and the system that converts that gap into a learning path automatically, so your L&D team can spend their time on program quality and manager enablement instead of manual path curation.
What we are building toward
We are early in this work. I will not pretend otherwise. What we have today closes the most expensive version of the gap — the onboarding problem, the new hire velocity problem, the "why is the 90-day ramp still 150 days" problem. That is the version that has the clearest ROI and the most immediate organizational pain.
The longer-term vision is a learning infrastructure where every gap-closure moment in an employee's career — not just onboarding, but every role change, every business evolution, every new skill domain the company needs — starts from the same gap-first logic. Where "we need to develop this capability in our team" produces a data-driven assessment of where the current population actually stands, and a path that closes the specific distance between where they are and where they need to be.
That vision requires more than a good onboarding tool. It requires L&D teams who are positioned to demand data-driven decisions from their organizations, and organizations that are willing to invest in the measurement infrastructure that makes evidence-based learning possible.
Building Learnforge is our contribution to that shift. Writing here is another one. If any of this resonates with what you are seeing in your own organization, I would genuinely like to hear about it.
— Jasmine Carter, CEO & Co-Founder, Learnforge | Nashville, TN