Independent evidence for technology decisions

Answer the hard questions your board is already asking about your technology.

Get an independent assessment of your systems, engineering, and technical risks.

Evidence-backed findings. Clear recommendations. Board-ready reporting.

Get a Read → See how it works

One conversation. No pitch deck. We'll tell you if we can't help.

Three weeks · Fixed price · Read-only · No engineering meetings

312 systems read → the decisions on your desk Kept current

Evidence of pressure Insufficient evidence

Every thread traced to a source record

From a real read · anonymized · Meridian Lending, a regulated consumer-lending platform See a real, redacted read →

Critical

Bus factor 1

One engineer authors 100% of production merges — and owns the auth core.

High

14% dead code

8,172 of 58,997 nodes built, shipped, and never used.

Critical

12/100

AI Act readiness — 8 consequential-decision systems, zero impact assessments.

One three-week read. Composite risk 8.2/10 · $1.4–4.8M annual exposure — every finding traced to its source record. Ask “prove it,” and the Read walks back to the commit.

Sound familiar?

Teams inherit systems. Leaders inherit narratives.

Critical decisions were made by people who have since left — no rationale, no trail. An independent read gives every hard question a clear answer you can prove.

01 →

“Who owns this?”

No one can say for certain — and the engineer who knew has moved on.

02 →

“What’s our AI exposure?”

You have a rough sense, but nothing you can put in front of a board.

03 →

“Are we on track?”

Delivery history lives across Jira, GitHub, and memory — unconnected.

04 →

“Did we just buy a liability?”

The diligence was surface-level; the real risk picture is still unclear.

05 →

“What did I inherit?”

You’re accountable for an estate you didn’t build — and the team that did has a stake in how it’s described.

06 →

“Will the code survive diligence?”

The picture so far is management’s narrative; the code itself hasn’t been read independently.

How it works

We connect the evidence into answers you can prove.

One read connects your plans, code, ownership, and delivery history — from your real records, never an AI’s guess — into answers you can prove.

Everything in one place

Plans, code, who owns what, and what’s shipped — connected in one place, not scattered across separate tools.

Answers leaders can use

Plain answers your leaders can take to the board — not raw metrics or a technical report.

Always current

Continuously maintained, so you can respond to hard questions as they arise — without commissioning a new engagement.

Fragmented evidence — plans, requirements, tickets, code, commits, comments, ownership, architecture, and delivery — grouped by semantic mapping into business domains such as auth, billing, platform, intake, AI surface, and risk, then surfaced as decision answers: roadmap drift, key-person risk, AI exposure, and diligence risk. — Fragmented evidence, grouped by semantic mapping, surfaced as answers you can act on.

Every answer traces back to source records. Ask “prove it,” and the read walks back to the commit.

Get a Read →

What you receive

Three weeks in. Here’s what arrives.

Not a 200-page PDF. A signed read your board can act on — every finding sourced to the evidence.

Executive briefing

The read in one signed document. Board-ready. Decisions surfaced, defended with sources.

Risk map

Active risks separated from dormant, technical debt quantified in dollars and time, ranked and mapped to your business domains. Sourced to the commit.

Knowledge concentration view

Who holds the keys, what breaks if they leave, where critical knowledge concentrates across your engineering organization.

AI / roadmap / architecture evidence layer

Your real AI footprint, your architecture as it actually is, and where the roadmap drifted from plan — each mapped to your domains and built for board reporting.

Kept-current evidence layer

The read stays live — re-queried, re-validated, and corrected as the system changes, so the next question starts from what’s true now.

Independence

We lead with the read. Independent. Defensible. Yours.

We don’t sell the rebuild, so the findings are not shaped by implementation work we hope to win.

Not a point-in-time audit

One-time audits are already stale when the next hard question arrives; this evidence base stays live.

Not a metrics dashboard

Delivery performance data doesn’t tell you why a decision was made or who is responsible for what.

Not dependent on people

Evidence is traced through artifacts — commits, tickets, documentation — not through whoever happens to still be around.

Who runs the read

Built by operators who’ve sat in your chair.

Hadar Wissotzky

Founding Partner

Serial founder and CTO — co-founded UrbanSitter, built and scaled multiple engineering orgs, and holds a US patent in business intelligence systems.

Peter Atkins