SPARK: How Coherent Solutions turned internal AI adoption into measurable enterprise value

The challenge

Coherent Solutions has a long track record of successful client delivery, but the specific details of that history were difficult to access. Information was spread across SharePoint folders, system fields, personal notes, and long email threads. Delivery managers carried years of context in their heads. Marketing teams rebuilt case studies manually, often without reliable sources. Sales leaders pulled fragments of information from multiple departments to prepare a proposal.

Across teams, this fragmentation increased effort, created inconsistency in branding and voice, and generated uncertainty about the reliability of sources in the organization.

Market pressure also added urgency to the project. Competitors over the past couple of years “rebranded” themselves as AI-Driven, AI-ready, and clients were asking more detailed questions about capability and delivery. Coherent had the experience to match those expectations, but it needed a structure that allowed teams to present its history clearly and consistently.

The problem was not a lack of data or capability, but rather the absence of a shared understanding and a reliable knowledge base.

Our approach

Coherent adopted a clear principle: AI should be embedded in the daily work of our thousands of professionals around the world, not added as discrete point solutions as an afterthought. This became the foundation for SPARK and the company’s broader efforts to modernize its organization and application of knowledge.

To put this principle into practice, SPARK followed a development model anchored by four key tenets:

1. Be fearless with innovation

SPARK began with a lightweight prototype that was released quickly to test its value. Early feedback showed that even an imperfect version of the platform addressed a real need.

2. Lead with learning

The team treated SPARK as a living system. Early versions were shared widely to uncover real use cases.

3. Build for adoption

SPARK had to deliver value immediately and remain reliable as it expanded. Widespread use was a primary goal because it autogenerated feedback to guide each iteration and kept the platform grounded in daily work.

4. Model what we advise

Taken together, these principles shaped SPARK as a system designed to remain unfinished. From the outset, the team treated it as a tool that improved through use rather than a product delivered once and optimized later. Feedback from real work plays a central role in how SPARK evolves, shaping what the platform prioritizes and how it responds to teams’ needs. Learning is continuous by design.

To successfully guide clients through AI adoption, Coherent needed an internal example of responsible use. SPARK is that example. The company’s commitment to embedding AI across its workforce extended beyond the platform and into how the organization structured adoption and governance around the technology. That commitment has also been demonstrated directly with prospects and clients, using SPARK to show how the platform operates in real delivery and governance contexts.

To support AI adoption across departments, the company created an AI Performance & Execution (APEX) governance model and an associated AI Champion network. Champions identify AI-related operational needs within their teams and translate those needs into practical use cases, many of which are implemented through SPARK. This keeps the SPARK product roadmap grounded in real work rather than abstract ambition.

The company also integrated SPARK into its internal digital ecosystem, including applications such as HubSpot, to reinforce sources of truth and to ensure data quality, reliability, and hygiene.

With the approach established, the effort shifted toward a solution that directly addressed meaningful opportunities to improve the daily work of employees.

The solution

SPARK began by addressing a specific need that was slowing down one operational process: The sales team needed to quickly and consistently produce accurate case studies. Rather than refining the manual process, Coherent built a system that made the underlying information easier to use.

SPARK unifies client, project, staffing, and marketing data to generate case studies, partnership histories, summaries, and analysis from governed, structured knowledge rather than individual memory. Outputs are generated using standardized formats and templates, ensuring consistency while allowing teams to quickly surface the most relevant context for each situation.

As SPARK matured, the team made a deliberate decision to anchor the platform to HubSpot as its system of record. While information existed across many of the organization’s tools, HubSpot contained the most structured and consistently maintained data on client relationships. Additionally, grounding SPARK in HubSpot ensured that AI outputs were based on governed, permissioned data rather than unstructured documents or unreliable sources. This decision marked SPARK’s transition from an experimental platform into a fully functional enterprise system.

Ultimately, integration with HubSpot enabled SPARK to reason over verified information, refresh automatically as teams completed their normal work, and avoid the risks associated with unsupervised data ingestion. Other accuracy safeguards were implemented by limiting data sources to approved systems and inheriting existing permission models.

SPARK also reduced knowledge loss by capturing client and project context through existing workflows, rather than informal memory. As teams document information in shared systems as part of their day-to-day work, that context remains accessible even when individuals change roles.

The team built SPARK using a flexible AI architecture that can evolve as the company’s needs change, with governance enforced through controlled data access, inherited permission models, and human oversight of AI outputs. These measures ensure SPARK remains safe, permissioned, and aligned with company values as its scope expands.

With the solution in place, the impact was immediately visible across teams.

The impact

As SPARK was widely implemented, teams were able to easily access:

• Case studies aligned with brand standards.

• Partnership journeys mapping the progression of client relationships.

• Account summaries built for QBR preparation.

• Custom capability overviews aligned to client business outcomes, industry or specific technologies.

• Staffing insights drawn from delivery history.

SPARK reshaped how teams prepare for, collaborate with, and support clients. Sales teams enter client conversations with specific success stories backed by clear documentation. Delivery managers consistently run QBRs with reliable context. The marketing team can quickly build assets from dependable sources.

With the adoption of SPARK, teams trust the outputs they receive. That trust is reinforced through clear boundaries and visibility into how the system is used. SPARK’s metrics are designed to support learning and governance rather than just optimizing to scale. Usage patterns help the team understand where the platform delivers value, where gaps exist, and how it should evolve next. By making AI behavior observable and grounded in approved data, SPARK supports confidence across teams rather than experimentation without oversight.

This has helped build confidence across the organization and shift preparation from reconstructing information to focusing on strategy.

There was also an equally important cultural shift in the organization. Team members who were uncertain about AI adoption now use SPARK for their everyday tasks. By giving SPARK concrete and immediate value, the platform fostered engagement organically.

The platform shows that Coherent can build AI with operational discipline, governed data usage, and human oversight of AI outputs. SPARK strengthens credibility with clients and reinforces confidence internally.

What’s next

Coherent designed SPARK as a platform that can continue learning new use cases without disrupting what already works. New capabilities can be introduced without reworking the foundation, while existing controls remain intact. This gives Coherent Solutions confidence to keep building on SPARK over time. The result is a system designed to evolve responsibly as AI becomes more deeply embedded across the organization.

Its structure also separates trusted information, AI reasoning, and outputs so each can evolve independently as needs change. This design allows the platform to grow deliberately across new teams and workflows while remaining grounded in existing governance practices. As more business functions across the company rely on SPARK for their value-driven use cases, this extensibility ensures the system scales through intent rather than accumulation.

SPARK’s ability to connect context across accounts, teams, and systems will continue to strengthen as relationships across the organization become more clearly defined.

The team is also exploring additional permission-based integrations that may eventually bring SPARK closer to daily communication channels, with explicit user control determining what information the system can access, including any future client-facing capabilities.

SPARK is not simply a tool; it is a system shaping how Coherent Solutions works in an AI-enabled environment.