How an AI WhatsApp Assistant Enrolled 2,337 Women Entrepreneurs in Colombia

9,220 women entrepreneurs contacted and 2,337 enrolled in just 25 days through WhatsAppIntelligent campaign segmentation was responsible for 80% of completed enrollmentsOnly 62 human interventions were needed to manage over 9,000 contacts45.6% of enrolled women needed a conversation with the AI to complete their registrationA team of 4 people achieved what would have required a call center of 15-20 agents

A foundation that transforms lives

Fundación Soy Oportunidad was born with a clear mission: to close the digital and employment gap for mothers living in poverty in Colombia. In a country where more than 8.5 million mothers raise their children under conditions of economic vulnerability, the foundation has become a bridge between these women and the opportunities the labor market offers — through digital education, entrepreneurship training, and workforce integration programs.

With programs like "Capitanas Emprendedoras" (Entrepreneurial Captains) and "Currículos Mothernos" (Modern Résumés), the foundation had spent years demonstrating that technology and education can transform realities. But nothing had prepared them for what would come at the end of January 2026.

Emprende Pro Mujer: closing gaps at a national scale

The Fondo Mujer Libre y Productiva (Free and Productive Women Fund) — an initiative of the Vice Presidency of the Republic of Colombia — in partnership with Pro Mujer, has been strengthening the capabilities of women entrepreneurs across the country. Over two years, the Emprende Pro Mujer program trained more than 4,000 women in all 32 departments of Colombia, delivered over 4,300 certificates, and developed a mentorship model that trained 76 leaders as entrepreneurship mentors.

The open call represented the next step: enrolling more than 2,337 new women entrepreneurs in 15 priority cities — Bogotá, Cali, Cartagena, Medellín, Santa Marta, Barranquilla, Pasto, San Andrés de Tumaco, Popayán, Armenia, Buenaventura, Pereira, Neiva, Florencia, and more — so they could be guided by mentors through a process of training and business development.

Fundación Soy Oportunidad was the organization tasked with making this call possible. And the task was no small feat.

The mirage of "free" software

In 2025, during Colombia Tech Week, the foundation had made an initial approach to TBit, the conversational AI agent platform. At that time, it was an experimental exercise — a demo where the foundation's captains connected participants with job opportunities through WhatsApp.

Shortly after, the foundation received what seemed like an ideal opportunity: free Salesforce licenses. For any organization, having access to one of the most powerful CRMs in the world sounds like the solution to every management problem. They hired an implementation partner and began the journey toward digitizing their processes.

What they discovered was a lesson that many organizations — both companies and nonprofits — learn the hard way: the cost of a technology tool is not just the license. It's the team that configures it, maintains it, adapts it, and operates it day after day. Implementation costs quickly exceeded the foundation's budget, and by the time it became clear that the Salesforce path wasn't viable, they were already racing against the clock.

The Fondo Mujer enrollment had non-negotiable deadlines. The number of women entrepreneurs who needed to register was also non-negotiable. And up to that point, the foundation had managed cohorts of at most 200 people.

Now they needed to scale from 200 to more than 2,000. In a matter of weeks.

A call at just the right moment

Sofía Valdivieso, Director of Operations at Fundación Soy Oportunidad, knew they needed a different solution — and fast. The foundation reconnected with TBit.

"We get it," recalls Clovis Rodríguez, co-founder of TBit. "We're a startup. We're not Salesforce or HubSpot. For a foundation with commitments to the Vice Presidency of Colombia, betting on a new tool can be nerve-wracking. But they gave themselves the chance."

The first days were about socialization: demonstrating that TBit wasn't simply a chatbot, but a complete platform capable of managing user attributes, understanding where each person stands in a flow, segmenting audiences, executing personalized mass campaigns, and — most importantly — doing all of this without needing an additional CRM.

In the final days of January 2026, Lucy was activated.

Meet Lucy

Lucy is a WhatsApp virtual assistant built on the TBit platform. But calling her a "chatbot" would be an understatement. Lucy was configured to be the first point of contact with the women entrepreneurs approaching the program — answering questions, guiding enrollments, resolving technical doubts, and accompanying each woman through her process, 24 hours a day, 7 days a week.

The strategy kicked off by combining Facebook and Instagram ads through Meta Ads with the WhatsApp channel. Interested women would see an ad about the Emprende Pro Mujer program, click through, and land directly in a conversation with Lucy.

And this is where the story gets interesting.

The happy path is the exception, not the rule

In any digital campaign, there's what's known as the "happy path" — the ideal user who sees the ad, clicks, completes the form, and registers without a hitch. In this enrollment drive, fewer than 20% of the women entrepreneurs took that path.

Colombia's reality is diverse and complex. There are women with low-end phones who struggle to load web forms. There are unstable internet connections in rural areas. There are entrepreneurs who had never interacted with a government digital enrollment process and had questions — Is this real? Do I qualify? What documents do I need? Is it safe to share my ID number on WhatsApp?

Lucy became the bridge between intention and action. When an entrepreneur got stuck midway through a form, Lucy knew exactly which step she had stopped at. When someone had questions about the requirements, they could ask in natural language and receive an immediate response. When the program's official form threw an error or gave confusing feedback, Lucy helped get the process back on track.

But the real differentiator wasn't the individual conversations. It was the campaign strategy.

The magic of intelligent segmentation

Over the 25 days of operation, the team executed 18 mass WhatsApp campaigns. But these weren't generic messages. Each campaign targeted a specific audience segment, built in real time from the information Lucy collected during every conversation:

• Already enrolled entrepreneurs: they didn't receive more enrollment messages. Respect for their time.
• Partial enrollments: personalized messages that included their name and exactly how many steps they had left. "Hi María, you completed 3 of 5 steps. Need help finishing?"
• Women who never started the process: re-engagement campaigns designed to address the most common barriers — fear, distrust, or simply lack of time.
• Facebook leads who never messaged on WhatsApp: a proactive first contact to connect them with Lucy.

This segmentation was responsible for 80% of completed enrollments. It wasn't a mass blast. It was a conversation at scale.

Sunday after Sunday

There's a part of this story that doesn't show up in any dashboard or metric: the dedication of the human team behind the operation.

Fundación Soy Oportunidad is a small team — no more than four people managing the entire enrollment operation. Sofía Valdivieso and her team worked practically without rest for the 25 days. Sundays included. Reviewing metrics, adjusting campaigns, identifying segments that needed attention, coordinating with the TBit team to optimize Lucy's flows.

Because technology doesn't operate on its own. Lucy could respond at 3 in the morning, but the strategy behind each campaign, the decision of which segment to prioritize each week, the analysis of which messages were driving the most conversions — that was done by a human team committed to ensuring every woman entrepreneur had her chance.

The TBit team supported the operation by providing technical support, strategic consulting, and platform adjustments to maximize every feature. But the operational credit belongs to the foundation.

The results

In 25 days of operation:

• 9,220 women entrepreneurs contacted through WhatsApp
• 14,666 responses generated by Lucy, the AI assistant
• 46,955 profile data points collected through natural conversations
• 18 mass WhatsApp campaigns executed with personalized segmentation
• 2,337 completed enrollments delivered to the Emprende Pro Mujer program
• 62 human agent interventions over the entire period — across more than 9,000 contacts

84.7% of the women entrepreneurs who completed their enrollment did so after receiving at least one proactive campaign. And 45.6% also needed a direct conversation with Lucy to complete the process — confirming that the AI's conversational capability is not an accessory, but a decisive conversion factor.

Only 14.5% of enrollments came through a direct path, without campaign or Lucy intervention. Meaning that without this operation, more than 1,900 women could have been left out of the program.

More than a form: the data Lucy collected

Lucy didn't just register entrepreneurs. In each conversation, naturally, she collected valuable information for the foundation and the program: email address, city, type of business, whether they had an active business, education level, annual sales, and even an 8-item self-efficacy scale.

She also identified special conditions of the participants — women who were victims of armed conflict, migrants, host community members, victims of gender-based violence, people with disabilities, peace signatories. Sensitive information that was collected naturally within the context of a conversation, not through a cold form.

These nearly 47,000 data points complement the program's formal survey and give the foundation a much richer view of who the entrepreneurs they are supporting really are.

"A turning point for growth"

"En alianza con TBit, decidimos hacer algo que para una fundación pequeña no siempre es fácil: profesionalizar la operación. Automatizar, ordenar, estructurar. Tomar decisiones estratégicas para poder crecer sin rompernos.

Gracias a esa decisión logramos gestionar la inscripción de miles de mujeres, organizar la información, dar respuestas oportunas y sostener el proceso con el rigor que ellas merecen.

No fue solo un reto operativo. Fue un punto de inflexión para el crecimiento.

Porque cuando una organización pequeña decide optimizar sus procesos, no está pensando en eficiencia… está pensando en impacto.

Y ese impacto hoy significa que más mujeres pudieron entrar al proceso, ser escuchadas y acercarse a una oportunidad real."

Translation: "In partnership with TBit, we decided to do something that isn't always easy for a small foundation: professionalize the operation. Automate, organize, structure. Make strategic decisions so we could grow without breaking.

Thanks to that decision, we managed to handle the enrollment of thousands of women, organize the information, provide timely responses, and sustain the process with the rigor they deserve.

It wasn't just an operational challenge. It was a turning point for growth.

Because when a small organization decides to optimize its processes, it's not thinking about efficiency… it's thinking about impact.

And that impact today means that more women were able to enter the process, be heard, and get closer to a real opportunity."

— Sofía Valdivieso, Director of Operations, Fundación Soy Oportunidad

What's next

Enrolling 2,337 women entrepreneurs was just the first chapter. Now the real program begins: months of training and mentorship, with 72 mentors distributed across cities throughout the country working directly with the participants.

For this next chapter, the foundation has activated Isla — a second AI agent on WhatsApp that is already connected with the program's mentors. As for what exactly Isla will do, we'd rather leave that as a preview for the second part of this story.

What we can share is that the vision is ambitious: giving a team of four people the ability to have real-time visibility into what's happening with thousands of participants distributed across the entire national territory.

Stay tuned.

This is the first part of a series on the use of conversational artificial intelligence in social impact programs in Colombia. If your organization faces similar challenges of scale in open calls, enrollments, or beneficiary follow-up, get in touch.