Customer story
How Speedinvest built the operating foundation for AI-native venture work.
As one of Europe's leading venture capital firms, Speedinvest had already seen the AI shift up close — founders, portfolio companies, and board conversations moving toward AI-native ways of working. Leadership saw the same opportunity internally: bring AI into the operating model of a regulated, operationally complex venture firm.
- Client
- Speedinvest
- Industry
- Venture capital
- Engagement
- AI enablement & systems integration
- Partner
- Specialty Tokens
Outline
From broad AI availability to practical AI capability.
Specialty Tokens partnered with Speedinvest to move from broad AI availability to practical AI capability — starting with company-wide enablement, moving into team-level workshops, and then focusing on the foundations required for AI-native venture operations: dealflow, compliance, CRM, ERP, email, meeting notes, and internal knowledge work.
Speedinvest had strong AI awareness and broad access to modern tools, but usage was uneven. Many high-value workflows still depended on systems of record, so a chat window beside the work was not enough.
Build shared AI fluency across the organisation, identify high-value team workflows, connect AI to operational systems, and support an internal AI lead who could continue the work beyond the engagement.
Speedinvest moved from company-wide AI access to a connected AI operating foundation: live workflows, AI-enabled teams, core systems connected to AI, and internal ownership to keep building.
The challenge
From AI access to operational AI.
Speedinvest did not need to be convinced that AI mattered. The firm already understood the shift through its work with founders, portfolio companies, LPs, and boards. But broad access did not automatically create an AI-native operating model.
Usage varied across teams. Some employees were already experimenting heavily, while others were still using AI mainly for drafting, summarising, or research. More importantly, many of the workflows with the highest potential value touched systems of record: ERP, CRM, compliance materials, email, meeting notes, and internal knowledge. The real opportunity was not simply better prompting — it was helping AI operate closer to the work itself.
The initial brief was to identify five to ten processes and automate them quickly. Once the engagement began, a more important foundation became clear: before Speedinvest could automate at scale, teams needed a shared understanding of what modern AI systems could do, how they could interact with business systems, and where human review still mattered.
Our approach
Build capability team by team.
Specialty Tokens started with company-wide AI upskilling and then worked with individual teams to translate the general opportunity into concrete workflows. The sessions were designed around real work, not abstract demos: teams explored where AI could help with their routine tasks, where expert review remained essential, and how repeated work could become reusable workflows.
That created a shared language across the organisation. Employees began to see AI not only as a tool for writing or answering questions, but as a system that could act inside external tools, retrieve operational context, and support repeatable business processes.
From the beginning, Specialty Tokens also worked closely with Speedinvest's internal AI lead. This was critical. The limiting factor was not only technical capability, but the time and ownership required to translate team processes into working AI workflows. By building internal ownership alongside the technical foundation, Speedinvest could continue the work after the engagement instead of depending on a one-off external project.
The foundation
Connect AI to systems of record.
The next step was technical: connecting AI to the systems that mattered. Speedinvest's ERP sat at the center of many operational workflows. Without a way for AI tools to interact with it, many automation ideas would have remained theoretical. Specialty Tokens began the ERP integration and extended the same approach to the CRM, working around an immature connector layer where necessary.
Once AI could interact with ERP, CRM, email, meeting notes, and related knowledge sources, teams could begin asking questions of operational systems, retrieving recent entries, and designing workflows that reflected the real state of the business. This moved AI from isolated prompts to connected workflows.
It also surfaced inconsistencies in the underlying systems of record. As workflows became more automated, data quality issues became easier to see and address.
Selected efficiency gains
Practical gains, with a foundation to keep improving.
Dealflow and CRM
Teams could work with company, pipeline, and relationship information more efficiently by connecting AI-supported workflows to CRM context instead of relying only on manual lookup and scattered notes.
ERP and fund operations
Operational questions that previously required manual navigation through systems could increasingly be supported through AI-assisted retrieval and workflow design.
Compliance and governance
AI-supported workflows helped teams find, summarise, and work with relevant documentation and process evidence more efficiently, while keeping human review in the loop.
Email, notes, and internal knowledge
Meeting notes, email context, and internal knowledge became more useful when connected into repeatable workflows rather than remaining isolated across tools.
Team-level workflow creation
The most important gain was not a single automation. It was the ability for teams to keep identifying, shaping, and improving their own AI-supported processes with support from an internal owner.
The real opportunity was not simply better prompting — it was helping AI operate closer to the work itself.
Then and now
Where the work moved.
Speedinvest had broad AI access, but usage was uneven. Value was often trapped in chat windows and prompts. Workflows were disconnected from systems of record, and AI ownership was still emerging.
Speedinvest has a connected AI operating foundation. AI can support workflows across ERP, CRM, email, notes, and internal knowledge. Teams have been upskilled, live workflows are in place, and an internal AI lead is running the work day to day.
Outcome
An operating foundation to keep building on.
The engagement moved Speedinvest from AI access toward AI-native operations — shared capability, connected systems, practical workflows, and internal ownership. The result is not a finished transformation. It is something more valuable: an operating foundation Speedinvest can keep building on, with the tools, team understanding, and internal ownership to keep rolling toward becoming a truly AI-native venture firm.