Tacentia
From tacit: the knowledge a firm never writes down.
We turn what your firm knows into technology you own.
The thing that makes your firm good is currently a liability. It lives in four partners’ heads. It can’t be sold, can’t be scaled, can’t be hired for, and it walks out of the building every evening.
Not an AI strategy. A working product, in weeks.
01
Why now
Everyone else in this market sells efficiency: save your team hours. That is a cost argument, and cost arguments lose to inertia. We are not selling hours back. We are turning the firm’s knowledge into something the firm owns.
What your firm knows is either an asset on your balance sheet or a risk in your succession plan, and right now it’s the second one.
02
Who this is for
Independent, owner-led US accounting and advisory firms, roughly $10M to $100M revenue, call it 60 to 400 people, where one of three things is true.
You have a method that could be a product.
A way of scoring, assessing, ranking or deciding that clients would pay for if it existed as software. An R&D credit approach, a valuation model, a succession-readiness framework, a niche advisory method. Today it’s a Word template and a partner’s instinct.
Your senior people spend their days on judgment-heavy grind.
File and workpaper review, transaction categorization against your rules, client onboarding and risk scoring, proposal scoping, anomaly flagging. Work that needs their judgment but not their hours. We extract the judgment and give the hours back.
You sit on data nobody can use.
If you have 300 construction clients, you hold the best construction benchmarking dataset in your state, and it currently drives nothing. Built out, it becomes a revenue line, a reason for prospects to call you, and a capability no competitor can buy.
The hard filter
No PE platform behind them. A platform-owned firm brings procurement, a group CTO and a nine-month cycle. Owner signs, or we pass.
Beyond that: the problem is specific, there is a clear first thing to build, and they can say what “worked” would look like.
No baseline, no build.
03
What we would build
Workpaper and file review assistant
First pass over workpapers against the firm’s own review checklist; flags exceptions with the reason and the source page; the reviewer signs off. Nothing is approved without a person.
Transaction categorization against the firm’s rules
Bank and ledger transactions categorized by the firm’s rules and its client-specific exceptions, with confidence shown and low-confidence items routed to a person.
Client onboarding and risk scoring
New client intake scored for fit, risk, complexity and expected margin at the door, with reasons, so the partner decides in minutes instead of discovering in month three.
Engagement margin and capacity overview
Realization per engagement while it runs, capacity by person, unbilled WIP, clients gone quiet, with alerts to the partner’s phone.
A niche advisory method as a scored product
An R&D credit approach, a valuation model, a succession-readiness framework or a cash-flow health check turned into a scored tool the firm’s clients use, under the firm’s brand. Today a Word template and a partner’s instinct; afterwards a product with a number on it.
The industry benchmark from one client base
300 construction clients is the best construction benchmarking dataset in the state. Anonymized, aggregated with the firm’s consent and the clients’ consent, published as a quarterly benchmark: a revenue line and a reason prospects call.
04
How we work
Define, Prove, Build Out. Every engagement starts small enough to judge us.
- Define1 to 2 weeks, fixed price
- Prove4 to 8 weeks, fixed price
- Build OutPhased or monthly
Define
Jordi diagnoses the problem, reviews what exists, and delivers a written plan: what to build first, what it costs, what “worked” means in measurable terms, what the data actually looks like, and what the firm has to supply. Includes a written data and confidentiality posture, and an independence check where the firm does attest work.
You can take that plan anywhere. Most people bring it back to us. The Define fee credits in full against Prove.
Prove
A working MVP on your real data, demoed weekly, live at the end. Not slides. The core proves itself before anything bigger gets funded.
Build Out
From MVP to full product: more features, more data, more users. In fixed phases, or as an ongoing partnership with defined capacity.
The working rules
The timeline risk sits with us. Fixed prices, because the deadline is fixed. Plan before code. Weekly working demos: not status reports, running software. Problems surfaced the same day we find them.
05
What you own
You own: the application, its configuration, your data, your method as expressed in it, your scoring logic, your brand, plus full source and documentation. Perpetual, irrevocable, no strings. You can hire anyone you like to maintain or extend it.
We retain: the underlying frameworks, libraries, architectural patterns, evaluation harnesses and generic components we use across engagements. These are our tools, not your product. You get a perpetual license to them as built into your system.
We never touch your data. No training on it. No aggregation. No reuse across clients. No exceptions without a separately signed agreement. In this profession we say that loudly and early, because everyone else is vague about it.
06
What happens when we’re gone
Every operator asks this, and they should. The answer is built into how we work, not bolted on at the end:
- You hold the source, the documentation and every access credential from day one.
- Subscriptions and infrastructure are in your firm’s name, on your firm’s card, from week one. You are never renting access from us.
- It is built so any competent developer can pick it up, and the handover pack says how.
- An ongoing partnership is available. It is a choice, never a dependency.
07
Where to start
Every engagement starts with a Define phase: one to two weeks, fixed price from $10,000, credited in full against the build. You get a written plan you can take anywhere.
08
Who you are dealing with
Peter Holgate
Peter opens the conversation and qualifies the fit. He has built six businesses from scratch and scaled them, creating over nine figures in enterprise value along the way, and has spent the last seven years finding and fixing the system problems inside owner-led businesses, most recently as co-founder of ClockWork League, a growth-systems program for agency owners. He sees the opportunity nobody else in the room sees, and says so.
Jordi Buskermolen
Jordi diagnoses the problem and builds the product. 25 years shipping software. Co-founded and ran a 30-person agency for 12 years (Cannes Lion, Webby). 18 months building AI products daily. AI writes most of the code; the decisions are his. That is why five weeks is realistic for one senior builder where a team would quote five months.
09
Proof
A brand agency’s proprietary index
A methodology turned into a scored, public product, live for their annual event, in five weeks, fixed price, on time.
BookmarksIQ
An AI SaaS with billing and a production MCP server, built and run solo. Proves real product, not prototypes.
A lead-screening instrument
A two-score AI research tool with deterministic scoring, in production. Proves the judgment-extraction claim specifically.
10
What we do not do
- AI strategy decks without building. We build or we don’t engage.
- Chatbots for the sake of chatbots, no-code platform assembly, “put AI on the website”.
- Enterprise vendor onboarding. If the deal needs procurement, SOC2 questionnaires and a nine-month cycle, it is the wrong deal for a two-person partnership. Owner signs, or we pass.
- Anything where “what would success look like” has no answer.
- Anything touching audit independence. Not in year one.
Talk to Peter
A 45-minute conversation. He will tell you whether this is worth a diagnosis.