Your process lives in someone's head
If your operations lead got hit by a bus, the business would forget how it works in 48 hours. Tribal knowledge is the foundation — and it's a fragile one. Nothing can be automated until it's documented.
Before AI,
you need foundations.
Most operations we walk into aren't AI-ready — and most don't know it yet. No documented process, three systems that don't talk, data scattered across six tools, no centralized source of truth. We fix that first. Sometimes the answer ends up being Zapier and a better SOP, not Claude. We say so. Foundations is honest pre-AI work — process mapping, systems audit, tool selection. It's where we start when 'AI for our business' is the goal but the foundation isn't there yet.
02 / Signals
Five signs you need
Foundations first.
These are the conversations where we say 'AI later — let's get your operation right first.' Two or more of these and Foundations is almost certainly the cheapest, fastest first step.
Your data is in six places
Customer info is split across the CRM, accounting system, email, a spreadsheet, an old Google Doc, and the founder's phone. There's no single answer to 'how many customers do we have?' AI can't see what nobody's gathered.
You've bought tools that didn't stick
There's a CRM you stopped using, a project management tool nobody opens, a Zapier account on auto-renew. The tools weren't the problem — the process around them was. Throwing AI on top won't help.
Every new hire onboards differently
There's no documented way to do anything, so each new person learns by osmosis from whoever's nearest. Six months later they're doing the work three different ways. AI would just codify the chaos faster.
You don't know what to automate first
You have a sense AI could help, but you can't say where exactly. That's not an AI problem — it's a visibility problem. Foundations gives you the map. Then we (or you) can decide what to ship.
03 / What we ship
Six pieces of work
we ship under Foundations.
Foundations isn't one thing — it's a menu of pre-AI work, scoped to where you actually need help. Most engagements use 2–4 of these. We pick what's actually missing, not what sounds thorough.
/01
Process Mapping
Half- to full-day shadow with the people doing the work. We document the actual process (not the documented one) — every step, every handoff, every workaround. Output: a real process map that matches reality, plus a list of friction points ranked by cost.
An example
Accounting firm discovered their 6-day month-end close was 22 distinct steps across 4 people, with 7 manual handoffs they didn't know existed. Map alone unlocked a 30% time cut before any tool was involved.
Typical inputs
- Time with the people doing the work
- Existing docs / SOPs (rough is fine)
- Access to the actual systems they use
- An honest leadership conversation about what's broken
- Visual process map (Lucid / Mermaid)
- Friction-points ranked report
- Quick-wins list (no tool required)
- Recommended next steps
/02
Systems Audit
We inventory every tool you're paying for, every subscription, every account. Map what each one is actually used for. Identify redundancies, gaps, and the tools that should be retired. Most audits find $500–$3K/mo of overlap that can disappear quietly.
An example
Services firm cut $14K/yr in SaaS subscriptions after the audit — three tools were doing the same job, two were unused, one was perfect for everything they'd been working around.
Typical inputs
- List of every subscription (we'll help build it)
- Login access (read-only) to current tools
- Department-level use cases
- Budget for replacement / consolidation
- Inventory + cost spreadsheet
- Tool consolidation plan
- Migration sequence
- Estimated savings + payback period
/03
Data Centralization Plan
Most SMBs have data — they just can't see it. We map where every important number lives, what shape it's in, and what it would take to bring it into one place AI (or any analytics tool) can later use. Sometimes the answer is a real warehouse. Often it's just getting three systems to talk.
An example
Property mgmt firm went from 'we have no data' to one canonical customer database in 3 weeks. Set the stage for the AI agent work that came next.
Typical inputs
- Map of current systems (from the Audit)
- What questions you want to answer
- Data privacy constraints
- Existing exports / API access
- Data flow diagram
- Centralization architecture
- Source-of-truth definitions per entity
- Integration sequence + estimates
/04
Tool Selection
When the answer isn't custom AI — and often it isn't — we help you pick the right off-the-shelf tool. Zapier vs. n8n vs. Make. HubSpot vs. Pipedrive vs. Close. QuickBooks vs. Xero. We've tried them. We'll tell you which fits your operation and which is a trap.
An example
Services business needed a CRM. We saved them 3 weeks of evaluation and a $400/mo wrong-tool subscription by pointing them straight at the one that fit. No AI involved. Bill: $1,800.
Typical inputs
- Use case + requirements (we'll help articulate)
- Budget range
- Existing tools to integrate with
- Team's tech comfort
- Shortlist of 2–3 tools with honest pros/cons
- Recommendation + rationale
- Setup + migration plan
- Optional: we do the setup
/05
Off-the-Shelf Integration
Sometimes you don't need custom — you just need the tools you already pay for to actually talk to each other. We wire up Zapier, n8n, native integrations, and webhooks. It's not glamorous AI work. It is often the highest-ROI 5 days of work we do.
An example
Brokerage connected their CRM, accounting tool, and DocuSign with 4 Zapier workflows. Eliminated 6 hrs/week of copy-paste. Cost: $3K. Payback: 4 weeks.
Typical inputs
- List of tools that need to talk
- What data should flow where
- Manual workarounds people do today
- Existing automation accounts
- Documented Zapier / n8n / native workflows
- Run-book for your team
- 30-day post-launch monitoring
- Optional: AaaS retainer for ongoing
/06
SOP & Documentation
The work that doesn't sound exciting but quietly unlocks everything else. We sit with your team, document what they actually do, and turn it into living SOPs — readable, maintained, easy to update. Often the first step to onboarding AI later, because you can't automate what nobody's written down.
An example
Construction firm documented 14 core SOPs in 3 weeks. New-hire ramp time fell from 3 months to 6 weeks. Half of those SOPs later became candidates for workflow automation — but only because they existed.
Typical inputs
- Time with your team
- Existing docs / training material (rough)
- Where docs live today (Notion, Drive, etc.)
- Who owns each process
- Documented SOPs in your wiki
- Visual checklists where useful
- Maintenance schedule + ownership
- Identification of automation candidates
04 / How it works
Foundations through
the Framework.
Every Sling engagement runs through the same five stages — Assess, Design, Build, Train, Optimize. Here's what that looks like specifically for Foundations. See the full Framework →
Listen, shadow, map
Half-day with leadership. Half- to full-day shadow with the people doing the work. Output: the rough map and the prioritized friction list. This is most of the value.
Pick the right moves
From the map, we decide what's worth doing now. Sometimes process. Sometimes a tool swap. Sometimes integration. Rarely (yet) AI. Output: a phased plan with cost + timeline.
Ship the foundation
Document the SOPs. Wire up the integrations. Migrate the tools. Whatever the plan calls for. Usually 1–3 weeks of focused work. Your team uses it on real workflow during build.
Make it stick
Document the new way. Train the team. Identify the in-house owner who'll maintain the SOPs. Without an owner, the new system rots back into the old one within 6 months.
And — only now — AI?
With the foundation in place, you can decide whether to keep going. Maybe the documented SOPs become Workflow Automation candidates. Maybe the centralized data feeds a Dashboard. Or maybe you're done, and that's a fine outcome.
05 / Toolbox
What we reach for
during Foundations.
Anthropic-first, tool-flexible. We pick what fits your data, your stack, and your team's appetite for new platforms. Nothing here is locked in — if you've already standardized on something else, we work with it.
Mapping
Lucid · Mermaid · Notion
Where we draw the process maps and document the SOPs. We pick what your team will actually keep up.
Integration
Zapier · n8n · Make
The no-code / low-code automation tools. Cheap, fast, often the right answer when the workflow is rules-not-judgment.
Off-the-shelf
HubSpot · Pipedrive · QuickBooks · Xero
The SaaS we know well enough to recommend honestly. We've used them, broken them, recovered them.
Data
Airbyte · Fivetran · Postgres
When centralization needs real plumbing. Off-the-shelf when possible, custom when necessary.
Docs / Wikis
Notion · Confluence · Google Drive
Where the SOPs live. We meet you where you are — we don't migrate you to Notion just because we like it.
Pre-AI honesty
Plain English
The most underrated tool. A well-written SOP often beats a half-built automation.
06 / Proof
One shipped example.
A representative engagement. Anonymized industry tag, real-shape metrics. The full case-study library lives at /case-studies.
30% time saved · no AI
Mapped a 6-day month-end close into 22 documented steps. Cut 30% before a single tool changed. Then identified the 4 steps worth automating later.
The shape
- Assess: full day shadowing the close team. Found 7 manual handoffs they didn't know existed. Three steps were being done twice. Two were vestigial — from a 2019 client who'd long since left.
- Design: cleaned-up process map, friction-points ranked. Recommended: kill 2 steps, merge 3, sequence 4 differently. No new tools.
- Build: documented the new SOPs in their Notion. Set up 2 Zapier workflows to eliminate the manual handoffs. 4 days of work.
- Train: 90-min team session. Identified the controller as the SOP owner. Maintenance cadence agreed.
- Optimize: 3 months later, came back for Workflow Automation on the 4 steps the map had flagged. By then they were ready — the foundation was solid.
07 / Pricing
Published. Honest.
Quoted in 48 hours.
We publish the range so you can scope before the call. The actual number lands inside that range — we quote it after Discovery, fixed-bid for Projects and MVP Sprints, hourly via AaaS for ongoing.
$2K – $4K
- Process map of one workflow
- Friction-points ranked report
- Quick-wins list (no tool required)
- Best fit for 'where do we start?'
$5K – $12K
- Full Assess → Train cycle
- Process map + systems audit + tool selection
- Integration work where needed
- Most common shape for this service
$500 – $1.5K / mo
- 2–8 hours per month, your call how to use them
- Lite or Standard tier
- SOP maintenance + tool monitoring
- Quarterly review + tune-up
- Pause anytime
08 / FAQ
Common questions
about this service.
What buyers actually ask on Discovery Calls about Foundations. If yours isn't here, we'll answer it on the call.
Because the alternative is paying us (or anyone else) 5x as much to build AI on top of an undocumented mess — and watching the AI codify the mess faster. The process map is the cheapest, fastest leverage point. It often produces 20–40% time savings before any tool is involved.
09 / Next stepShow us the
Show us the
process you can't
actually describe.
The one your team does differently every time. The one nobody can explain in five sentences. Bring it. We'll spend 30 minutes on it. If Foundations is the right first step, you'll get a scoped quote in 48 hours. If you're already foundation-ready, we'll route you to the AI service that fits.
What this call is
30 min. We learn your operation, scope the work, tell you straight whether Foundations fits.
What you'll leave with
A scoped quote within 48 hours, or an honest "this isn't us" and a referral to a better fit.
Other services
Foundations almost always leads somewhere — usually Workflow Automation, sometimes Dashboards or Apps. We sequence the work honestly.