Customer Support

How to build a customer knowledge base as Professional Services Founders

Customer SupportFor Professional Services Founders2 apps11 steps~22 min to set up

Your consultancy's institutional knowledge lives in four places simultaneously: a Notion wiki nobody updates, a Google Drive folder with 200 files named 'v2_FINAL_USE THIS', a Slack thread from six months ago, and your own head. A new associate spends their first two weeks asking you the same questions your last three hires asked. Client-specific context — their procurement quirks, the stakeholder who needs a pre-call, the scope creep clause that bit you in Q3 — gets rediscovered the hard way every engagement. You know you should document it. You also have three proposals due and a client QBR on Thursday.

Build this on Starch → See the apps used

Customer SupportFor Professional Services Founders2 apps11 steps~22 min to set up

Outcome

What you'll set up

A single searchable knowledge base that pulls from your existing Notion pages, Google Drive docs, and Gmail threads — so associates find answers without interrupting you

An AI layer that auto-categorizes new content, flags when a client brief or methodology doc goes stale, and builds role-specific onboarding paths for new hires

A client-context hub per engagement — procurement contacts, scope history, key decisions, past deliverables — surfaced in one place before anyone opens a proposal doc

The Starch recipe

Apps, data, and prompts

The combination of Starch apps, the data sources they pull from, and the prompts you use to drive them.

Apps used

Knowledge Management Email Agent

Data sources & config

Starch connects directly to Notion (scheduled sync — pages and databases refresh on a schedule); connects Google Drive from Starch's integration catalog, queried live when a search or summary runs; and syncs your Gmail on a schedule to pull client threads and surface decisions. Slack is also available from Starch's integration catalog for querying channel history live. No browser automation required for this stack — all connections are API-based.

Prompts to copy

Build me a client knowledge base that pulls from our Notion workspace and Google Drive. For each active client, create a page that shows: key contacts and their communication preferences, scope boundaries from the SOW, decisions made this engagement, and any known landmines from prior work. Make it searchable by client name, project type, and industry.

Set up an onboarding path for new associates that surfaces the ten most important internal methodology docs, our standard deliverable templates, and a FAQ section I can populate. Flag any doc that hasn't been edited in 90 days as potentially stale.

Connect my Gmail so that when a client thread contains a decision, a scope change, or a new stakeholder introduction, it gets logged to the right client page automatically.

Run these in Starch → or paste them into your favorite agent

Walkthrough

Step-by-step

1 Connect Notion as a scheduled-sync provider — Starch pulls your existing pages and databases automatically, so the knowledge base starts with what you already have rather than a blank slate.

2 Connect Google Drive from Starch's integration catalog; the agent queries it live when a search runs across your deliverable templates, past proposals, and methodology docs.

3 Sync Gmail through Starch's scheduled-sync connection — the agent reads threads (30 messages per page) and extracts client decisions, new stakeholder introductions, and scope-related language.

4 Tell Starch to build a client knowledge base: describe each client page should include key contacts, communication preferences, SOW scope summary, open risks, and a log of decisions made this engagement.

5 Tell Starch to auto-categorize incoming Notion pages and Drive docs by client name and project type — so new files filed by anyone on the team land in the right place without manual tagging.

6 Set a staleness rule: any client-context doc or methodology page not edited in 90 days gets flagged for review. Starch surfaces these in a weekly digest so you're not caught off-guard when a client asks about a process you updated a year ago.

7 Build an associate onboarding path — a structured sequence of pages Starch surfaces to new hires in their first two weeks, including methodology primers, deliverable standards, and client etiquette notes.

8 Use the Email Agent app to triage your Gmail inbox by actual client priority, summarize long threads, and draft replies — so that when a client emails with a question that's already answered in the knowledge base, you can respond in one click with the right context pulled automatically.

9 Set up a prompt that generates a pre-engagement briefing doc for any active client — Starch compiles key contacts, prior decisions, scope limits, and known risks into a one-page brief your team reads before a kickoff or QBR.

10 Publish the knowledge base to your team inside Starch — associates search it directly, reducing the volume of 'quick questions' that land in your Slack DMs mid-proposal sprint.

11 Review the staleness queue monthly and use Starch to draft updated versions of flagged docs based on what's changed in recent Gmail threads and Notion edits — so documentation stays current without a dedicated documentation sprint.

See this running on Starch

Connect your tools, describe what you want, and the agent builds it. Closed beta is free.

Try it on Starch →

Worked example

Northgate Advisory onboarding a new senior consultant, March 2026

Sample numbers from a real run

Client pages auto-populated at setup	14
Methodology docs indexed from Google Drive	47
Stale docs flagged in first staleness scan	11
Associate onboarding questions answered without founder involvement	23
Hours saved in first two weeks vs. prior hire onboarding	8

When Northgate Advisory brought on a senior consultant in March 2026, the founder connected Notion (scheduled sync) and Google Drive (live query from Starch's integration catalog) and described the knowledge base in plain language. Starch indexed 47 methodology and deliverable docs from Drive and built 14 client pages from existing Notion content, each showing the key contact list, scope summary, and a decision log pulled from Gmail threads. The staleness scan flagged 11 docs — three of which were client process guides that had been quietly superseded. In the new hire's first two weeks, the associate asked 23 questions that the knowledge base answered directly. The founder tracked the comparison to the prior hire: that cohort generated roughly 30 interruptions in the same window, averaging 20 minutes each. The knowledge base didn't eliminate all questions, but it handled the ones with documented answers — which turned out to be most of them.

Measurement

How you'll know it's working

Associate ramp time to first unsupervised client deliverable (target: under 3 weeks)

Volume of repeat questions escalated to the founder per month (goal: below 5)

Percentage of active client pages with a decision log updated in the last 30 days

Number of stale docs flagged and updated per quarter

Pre-engagement brief completion rate before kickoff calls (are associates actually using it)

Comparison

What this replaces

The other ways teams handle this today, and how the Starch version compares.

Notion alone

Great for structured pages but has no AI search layer, no staleness detection, and no Gmail integration — so it stays current only if someone manually updates it, which they won't.

Guru or Tettra (dedicated knowledge base tools)

Purpose-built for this, but they sit outside your existing HubSpot/Notion/Gmail stack and require content to be migrated and maintained in a separate system — adding a tool rather than connecting the ones you have.

Confluence

Powerful for large engineering orgs, but for a 12-person consultancy the setup overhead is disproportionate and it doesn't connect to Gmail or Notion without custom integrations.

Google Drive search + manual naming conventions

Free and already in use, but search quality degrades fast past 200 docs, there's no AI extraction of decisions from Gmail, and new hires have no structured path through the content.

On Starch RECOMMENDED

One platform — knowledge management, email agent all running on connected data. Setup in plain English; numbers stay current via scheduled syncs and live agent queries.

Try it on Starch →

FAQ

Frequently asked questions

We already have a Notion wiki. Will Starch just duplicate it?

No — Starch syncs your existing Notion pages on a schedule and makes them searchable through the knowledge base. You're not migrating content; you're adding an AI search layer and a Gmail connection on top of what you already built. Pages you update in Notion show up in Starch on the next sync cycle.

Can Starch read client emails and log decisions automatically, or do we have to tag things manually?

Starch syncs your Gmail on a schedule and the agent reads threads to extract decisions, scope changes, and new stakeholder introductions. You describe the extraction rule in plain language — something like 'if a thread mentions a scope change or a decision, log it to the matching client page' — and Starch applies it going forward. It's not perfect on ambiguous threads, but for clear decision language it's reliable.

Is Starch SOC 2 certified? We work with enterprise clients who ask about data security.

Not yet — Starch is not SOC 2 Type II certified. If your enterprise clients require a SOC 2 report as a condition of tool use, that's a real constraint worth naming upfront. It's on the roadmap.

What happens to Google Drive files that aren't in Notion — do they get indexed?

Yes. Google Drive connects from Starch's integration catalog and the agent queries it live when a search runs — so Docs, Sheets, and Slides in Drive are searchable alongside your Notion pages. The Drive connection doesn't require content to be in Notion first.

What about the Customer Support Agent app — can we use that for client-facing FAQs?

Customer Support Agent is currently in development — it's coming soon and you can request beta access, but it's not available today. For now, the Knowledge Management app handles internal search and the Email Agent handles inbox triage. When Customer Support Agent launches, it will use your knowledge base as its source of truth for client-facing questions.

Can we give clients read access to a subset of the knowledge base — like a client portal?

Not today in a formal client-portal form. The knowledge base is currently an internal tool for your team. You can describe a custom surface to Starch — something like 'build me a client-facing FAQ page for each active engagement, drawn from this client's knowledge base page' — and Starch can build a view of that content, but access management for external users is not a built-in feature yet.

How does the staleness detection actually work — will it nag us constantly?

You set the threshold when you describe the app. A reasonable starting point for a consultancy is 90 days for methodology docs and 30 days for active client pages. Starch surfaces flagged docs in a scheduled digest — weekly or monthly, your call — rather than pinging you every time something ages out. You review the list, update the ones that matter, and archive the ones that don't.