Massively AI

Fractional CTO Support for Scalable Growth

Massively, an early chatbot company, was down to two mid level developers after its engineering team moved on, with no one left to modernize its stack. Ollon served as fractional CTO for five years, moving the platform off AWS OpsWorks to Docker and integrating new LLMs into the chatbot pipeline. Massively kept shipping through multiple rounds of turnover, staying current as AI shifted.

Situation

Massively creates and customizes chatbots for companies with online sales platforms, seamlessly integrating with live agents to support customers. As the company was established when chatbots were newer and more novel, they had a relatively large team that had scaled down. The original team that built the engine had moved on, leaving the company with two mid level developers.

Solution

Ollon helped Massively transition to newer technologies, implement modern DevOps practices, and mentor the development team. As large language models (LLMs) like ChatGPT emerged, we guided the integration of these tools into Massively’s existing chatbot infrastructure, ensuring their platform remained current, competitive, and aligned with evolving customer expectations. Additionally, we supported spikes in work, bridging gaps with our developers, and managed hiring as needed over five years of working with them.

Results

Working within a modest budget, Ollon acted as a fractional CTO, providing architectural leadership, staff augmentation, and continuity through team turnover. The addition of LLM integration positioned Massively to stay relevant in a rapidly changing AI landscape while maintaining development momentum and controlling operational costs.

Our Expertise in Action

AI Integration

An AI focused startup's chatbot engine predated the current wave of large language models, and its two remaining developers had no plan for folding tools like ChatGPT into a pipeline built before those models existed. Ollon decided against replacing the engine outright, integrating new LLMs directly into the existing pipeline, feeding each model from the company's own structured knowledge source and formatting its responses for the customer. That integration point became the seam between the startup's own data and whichever LLM was current, so a newer model could be swapped in without touching the rest of the pipeline. The chatbot now answers customers using current LLM technology on the same infrastructure the company already depended on.

Fractional CTO and fractional technical leadership

An AI focused startup building customer service chatbots for companies with online sales platforms had lost the technical leadership of its original engineering team, leaving two mid level developers to make architecture calls on a budget that ruled out a full time executive hire. Ollon filled that role as fractional CTO across the multi year engagement, making calls like weighing a full rebuild against modernizing in place, and deciding how new AI models should get folded into the existing system instead of replacing it outright. That leadership gave the company consistent technical direction across five years and multiple rounds of team turnover that a startup at that budget level could not otherwise access.

Team mentoring and coaching

An AI focused startup building customer service chatbots for companies with online sales platforms was down to two mid level developers after its original engineering team moved on. Ollon's developer paired daily with the company's remaining internal developer, splitting the chatbot pipeline into two connected halves, with the internal developer structuring the underlying knowledge source while Ollon's developer fed that knowledge to the AI model and formatted its responses for users. The two halves were directly linked, so they reviewed each other's code constantly to keep the pipeline working end to end. This daily pairing kept chatbot development moving with just two developers on the team.

Team augmentation and staff scaling

An AI focused startup building customer service chatbots for companies with online sales platforms worked with Ollon over a multi year engagement, well beyond a single fixed project. When work spiked past what the company's own two developers could handle, Ollon added developers to bridge the gap, then scaled back down once the spike passed. Ollon also managed hiring on the company's behalf as its needs changed over time. This let the company handle uneven workload without carrying a larger permanent team than its modest budget could support.

Legacy system modernization

An AI focused startup building customer service chatbots for companies with online sales platforms had built its chatbot engine when the category was still new, and the team that originally built it had since moved on, leaving two mid level developers to maintain a stack that had aged around them, including a deployment model still running on AWS OpsWorks. Over a multi year engagement, Ollon helped transition that engine off OpsWorks and onto Docker, working through the platform section by section without pausing product work already in flight. The result was a chatbot engine running on a modern container based deployment, positioned to keep pace with a fast changing AI landscape.