An AI-Powered Knowledge & Sales System for BK Consulting
How BK Consulting replaced HubSpot, Fathom, and scattered Slack threads with one AI-connected Notion workspace that runs knowledge management, sales, and delivery end to end.
Three tools, none of them agreeing.
Commercial data sat in HubSpot, call recordings in Fathom, and the day-to-day conversation in Slack. No single tool was the problem. Pulling an answer out of all three took long enough that people stopped bothering and asked someone instead.
Storage was never the problem here. Three tools already stored plenty.
What was missing was one space where teams could work without stepping on each other, and where information moved between them without someone carrying it.
Centralize, connect, automate.
In that order. Automating a process nobody has agreed on just makes it faster to get wrong.
Centralize
One workspace where knowledge, pipeline, and delivery live together, searchable by AI.
Connect
Team spaces so Sales, delivery, and leadership share the same data without stepping on each other.
Automate
Agents handle the handoffs between stages so nothing waits on a person to notice and act.
What got built.
A searchable knowledge base
The workspace starts with a company homepage carrying everything an employee needs to know. On top of it, Notion AI searches the workspace and every connected source at once. Ask it something like what the corporate travel policy is, and it finds the answer directly instead of sending someone digging through folders.
A Sales system built on Notion, not HubSpot
The Sales team works out of its own team space: connected databases for accounts, the opportunity pipeline, and the services catalog, all with values attached. Everyone on the team sees the same data without needing separate access to a separate tool.
AI meeting intelligence
Every meeting starts inside a notes page. Notion AI transcribes live, distinguishing between speakers, while notes and an agenda get written alongside it in real time. Once the meeting ends, AI reads the transcript and produces a summary and action items, tailored to the kind of meeting it was, a customer call, a stand-up, a candidate interview, or a custom instruction. Each note attaches to the account or opportunity it belongs to.
Agents that move deals forward without anyone touching them
Moving an opportunity to Proposal triggers an agent that drafts a quote from the opportunity’s services and values. A second automation turns that quote into a PDF and sends it to the customer, running in the background while the rest of the team keeps working.
When a deal is won, another agent takes over. It creates a project for delivery and assigns it to a consultant based on region, specialization, and current capacity, pulled from the consultant directory. The opportunity’s full history, meeting notes and all, becomes the project brief automatically.
The opportunity becomes a living record of the project, so the handoff document writes itself.
One dashboard for the whole business
The same databases that power collaboration also power a management dashboard: business status, consultant capacity, pipeline forecast, and which opportunities need attention today, all in one glance, without opening a second platform.
What changed day to day.
No more tool-hopping
Knowledge, pipeline, and delivery information all live in one AI-searchable workspace.
Proposals go out without anyone drafting them
Moving a deal to Proposal triggers quote creation, PDF generation, and delivery automatically.
Handoffs happen with zero manual documentation
Won deals become fully-briefed consultant projects, assigned by region, specialization, and capacity.
Leadership sees the business in one glance
A single dashboard, powered by the same underlying data, replaces checking multiple platforms.
What I did on it.
This system runs BK Consulting itself. I designed and built it end to end, responsible for:
One set of records, three jobs.
Knowledge, pipeline and delivery share one connected space, which is what carries the weight here rather than any single automation. Nothing drops between a first conversation and a finished project.
AI does the heavy lifting where it helps: reading a transcript, matching a consultant to a project, drafting a quote. It stays out of the way everywhere else.
One set of records doing three jobs, with AI only where it earns its place.
If this sounds like your team, the same shape adapts. It starts with an audit of how your work moves now.