Silmari

Institutional memory for systems that outlive their teams.

FOUNDERMaceo Jourdan · Founder & CEO

EMAILmaceo.jourdan@gmail.com

THE PROBLEM

The people who know why are leaving. The emails that know are unreadable.

One live system. 13 years in production. 14 years of decisions.

200,000+

emails

change requests

the client lost

tickets

ServiceNow

docs

▼

Why is this on-prem and not AWS?

new CIO

What year did we go to Oracle 19c?

nobody knows

Can we turn off server 15?

nobody dares

THE PEOPLE

The project manager of 14 years leaves at year end. Her ~100,000 emails leave with her.

THE COST

A major city still runs a system that went live in 1998. Running it costs millions a year. Replacing it costs tens of millions, because no one knows how it was built.

THE SERVER

Someone switched off a forgotten lab server eight years after its builder left. It was sending every text message for an application serving tens of thousands of people.

VALUE PROPOSITION

Answer in minutes what now takes days of email archaeology — before the people leave, not after.

MOMENT

TODAY

WITH SILMARI

New CIO asks "why on-prem?"

Search years of email, ask whoever is left

Cited answer, with the original decision thread

Auditor asks why a transaction was booked that way

Ask 10 people, search inboxes

The decision and who made it, sourced

Key person resigns

Their memory walks out

Their archive is already memory

Legacy rewrite or migration

Months of research before the first line of code

The requirements and trade-offs, recovered and dated

THE VALUE

A $[[100M]] system you can still change, explain and replace.

[[Hours saved per question × questions per year, or replacement-cost avoided — needs one real number from the pilot.]]

UNDERLYING MAGIC

Derived memory: from raw archive to dated, cited facts you can ask about.

Slide 1 · Title

Open with one sentence

"I support a system that went live in 2013. I have over 200,000 emails about it, and every week someone asks a question only those emails can answer."

Slide 2 · Problem / Opportunity

Stories

All three stories are first-hand, from video 5VdmZIm6UfU. Tell one, not three, out loud. The pulse-rate story is the server.

Slide 3 · Value Proposition

No speaker note in the v14 draft.

Slide 4 · Underlying Magic

Switch to the demo here

[[Demo: load a real (sanitized) project archive and ask the Oracle 19c question live.]]

Slide 5 · Business Model

No speaker note in the v14 draft.

Slide 6 · Go-to-Market Plan

No speaker note in the v14 draft.

Slide 7 · Competitive Analysis

Source: research/2026-10-04_tam-sam-som-institutional-memory.md, section 4. Cited fact: Microsoft Learn (Exchange hybrid and Copilot, updated 2026-08-18) says Copilot is cloud-based and cannot ground on on-premises mailboxes. Glean's on-prem support is unknown, and no vendor with Oracle support was found. That is an absence of evidence, not proof about roadmaps.

Silmari's own cells are honest: Postgres works today, the Oracle port is in progress, and "minutes, cited" is a pilot claim not yet independently verified. Re-check Sentra and Copilot-for-Exchange-Server status right before the meeting.

Detail: Appendix A2.

Slide 8 · Management Team

No speaker note in the v14 draft.

Slide 9 · Financial Projections & Key Metrics

Assumptions

Every input is an assumption until paying customers exist. The $60K per system per year price is founder-set. Direct quota ($859K to $1.2M of new license ARR per account executive) is founder-set. Partner pilot-to-paid (35%, 40%, 45%) and direct win rate (25% for account executives, 30% for the founder) are assumptions. Model counts are fractional and rounded on the slide.

Market math

TAM: 137,522 US firms with 75 to 499 employees × 40% thin or outsourced IT = 55,009 firms × 2 systems = 110,018 systems × $60,000 = $6.60B a year. SAM: 55,009 × 50% with a system live 10+ years × 50% with a retained, exportable archive reachable by a partner = 13,752 firms × 2 = 27,504 systems × $60,000 = $1.65B a year. Year-3 systems: 211 / 27,504 = 0.77% of SAM systems. The 40%, 50%, 50% and 2-systems shares are assumed, not sourced. The earlier "hundred billion" headline figure was dropped as unsupported.

Throughput view

Throughput T = revenue − truly variable costs (partner share, hosting, commissions). Operating Expense OE = all salaries and other operating spend. Net Profit = T − OE. Year 3 is nearly break-even (T / OE = 0.92). The base case never runs out of cash: lowest cash is $726,684 in month 25 on the $3M raise. At a $36K price the model runs out of cash in month 23. Detail: Appendix A1, A4–A7 (A6 financial model, A7 throughput and unit economics).

Slide 10 · Status, Timeline & Use of Funds

No speaker note in the v14 draft.

Silmari Pitch Deck v14 — Institutional Memory (plain-text mirror)

Every system that has run for a decade has a second system behind it: the reasons. Why it's on-prem. When Oracle was upgraded. What that server in the lab does. Those reasons live in 200,000 emails and in the heads of people who are about to leave. Silmari turns that archive into memory you can ask questions of.

Canonical source for this document: silmari-memory_pitch-deck_v14-institutional-memory-DRAFT.md. Items shown as [[ ]] are unfilled placeholders.

Slide 1 — Title

Silmari. Institutional memory for systems that outlive their teams.

Maceo Jourdan, Founder & CEO. maceo.jourdan@gmail.com.

Slide 2 — Problem / Opportunity

The people who know why are leaving. The emails that know are unreadable.

One live system. 13 years in production. 14 years of decisions. Sources: 200,000+ emails, change requests the client lost, tickets / ServiceNow, docs. Questions nobody can answer: "Why is this on-prem and not AWS?" (new CIO). "What year did we go to Oracle 19c?" (nobody knows). "Can we turn off server 15?" (nobody dares).

The project manager of 14 years leaves at year end. Her ~100,000 emails leave with her.
A major city still runs a system that went live in 1998. Running it costs millions a year. Replacing it costs tens of millions, because no one knows how it was built.
Someone switched off a forgotten lab server eight years after its builder left. It was sending every text message for an application serving tens of thousands of people.

Slide 3 — Value Proposition

Answer in minutes what now takes days of email archaeology — before the people leave, not after.

Moment	Today	With Silmari
New CIO asks "why on-prem?"	Search years of email, ask whoever is left	Cited answer, with the original decision thread
Auditor asks why a transaction was booked that way	Ask 10 people, search inboxes	The decision and who made it, sourced
Key person resigns	Their memory walks out	Their archive is already memory
Legacy rewrite or migration	Months of research before the first line of code	The requirements and trade-offs, recovered and dated

The value: a $[[100M]] system you can still change, explain and replace. [[Hours saved per question × questions per year, or replacement-cost avoided — needs one real number from the pilot.]]

Slide 4 — Underlying Magic

Derived memory: from raw archive to dated, cited facts you can ask about.

Archive (emails, tickets, change requests, docs, code) to derive (who, what, when, why; facts linked to the source message they came from) to ask ("Who upgraded server 7 in 2018?", "Why is server 15 offline?"; answer plus the emails it came from).

Temporal. People are bad at years. Silmari dates every decision from the record, not from anyone's recall.
Episodic. It keeps the event and the reason: why the servers were upgraded, how, and by whom.
Durable. A 14-year memory can't live on a laptop. It runs on the databases enterprises already trust: Postgres today, Oracle port in progress, SQL Server [[and DB2]] if clients require.

Switch to the demo here. [[Demo: load a real (sanitized) project archive and ask the Oracle 19c question live.]]

Slide 5 — Business Model

Who has the money today: the budget that already pays for this problem, badly. [[Legacy-system run budgets, migration/modernization projects, audit preparation, and consultants hired to reconstruct how a system works.]]

Per-system / per-archive license — priced against the cost of not knowing, not per seat. [[$ per system per year]]
Ingestion engagement — a fixed-fee project to load and derive a system's history. [[$ per engagement]]
On-prem / private-cloud deployment for clients who can't send email off-site.

Slide 6 — Go-to-Market Plan

Start where the archive already exists and the departure is already scheduled.

Our own systems first. The founder's 13-year system is pilot zero: real archive, real questions, real departure date.
Systems integrators and long-term support contractors. They carry decades of client email and lose staff constantly. One partner opens many systems.
Database channel. [[Oracle ISV Partner Program — status from v13 was "in talks"; confirm before using.]]
Trigger events we sell into: key-person departure, new CIO/CTO, audit season, modernization RFP, cloud repatriation.

[[Cost to acquire one customer, and how long a sale takes.]]

Slide 7 — Competitive Analysis

	Enterprise search / Copilot over email	Knowledge bases (Confluence, Notion)	eDiscovery / email archive	Consultants doing archaeology	Silmari
Works on the archive you already have	Partly — needs the mail in the vendor's cloud or via connectors	No — someone must write it	Yes	Yes	Yes
Answers why, with dates	Finds messages, not decisions	Only if written down	Cited Q&A, built for legal review	Yes, slowly	Yes, cited
Survives staff turnover	Yes	Yes	Yes	No	Yes
Runs on Oracle / on-prem	Copilot: no. Glean: unknown	Mostly cloud	Mostly cloud; Cohesity Gaia: on-prem backups	n/a	Postgres today, Oracle port in progress
Time to answer	Minutes, unverified	n/a	Minutes to hours, unverified	Weeks	Minutes, cited (pilot claim)

Microsoft: Copilot is cloud-based and has no access to on-premises mailboxes. No vendor with Oracle support found. Closest: Glean, Microsoft 365 Copilot, Sentra. Detail: Appendix A2 (appendix/a2-competitive-landscape.html).

Slide 8 — Management Team

[[Carry the v12 team slide — Maceo / Landon / Matt Richter — and rewrite each line around institutional memory.]]

Maceo Jourdan, Founder & CEO — has supported the same production system since 2013, and built systems in the late 1990s that are still running. He is the customer.
[[Landon]] — [[role and why it matters for selling into enterprise]]
[[Matt Richter]] — [[role]]
Advisors / investors: [[ ]]

Slide 9 — Financial Projections & Key Metrics

Bottom-up, three years, $60K per system per year (founder-set). Year 1 / Year 2 / Year 3:

Systems under memory (year end): 9 / 63 / 211
New paid systems, direct: 6 / 36 / 75
New paid systems, partner: 3 (plus 1 founder pilot) / 21 / 86
Direct win rate: 25% each year. Partner pilot to paid: 35% / 40% / 45%
ARR (year end): $0.55M / $3.79M / $12.66M
Throughput (T = revenue − truly variable costs): $0.27M / $2.07M / $8.23M
Operating Expense (OE, all salaries and operating spend): $1.43M / $4.63M / $8.96M
Net Profit (T − OE): −$1.16M / −$2.56M / −$0.73M

Market, bottom-up at $60K per system: 55,009 thin-IT US firms × 2 systems = $6.6B TAM; 13,752 firms, 27,504 systems = $1.65B SAM; year-3 plan of 211 systems is 0.8% of SAM. Base case never runs out of cash: lowest cash $0.73M in month 25. Price sensitivity in A6. Detail: Appendix A1, A4–A7.

Slide 10 — Status, Timeline & Use of Funds

Where it is: [[confirm current state — v13 listed "alpha live, 15 MCP tools"; the database is built on Postgres with an Oracle port underway.]]

Next version: load a full 13-year project archive (200K+ emails) and answer the real questions that come in each week. [[Target date.]]

The raise: $[[3M]] · [[instrument]] · [[runway]]

Ingestion + derivation engine at 200K-email scale — [[ ]]
Oracle / SQL Server enterprise ports — [[ ]]
First [[N]] paid systems through integrator partners — [[ ]]
Security / on-prem certification — [[ ]]

Every time someone leaves, the system forgets a little. We're making sure it doesn't.