Case Studies

Three anonymized consulting engagements. Each one names the situation, the diagnostic, the intervention, and the outcome. No slideware, no named clients, no inflated metrics.

Request a Discovery Call View Technical Case Studies

Discretion policy

All engagements are anonymized. Industry, scale, and engagement size are real. Client names are never disclosed, even with permission. Case studies are written to be sharable internally without legal review.

HoloOS

Mid-cap fintech

Engagement
Organizational diagnostic + strategy
Size
$18k
Duration
6 weeks

Markers

  • Series B
  • $40M ARR
  • 120 engineers
  • 6-week engagement

Context

A Series B fintech (120 engineers, $40M ARR) had been running a "digital transformation" initiative for 18 months with sub-15% adoption across target teams. The COO was under board pressure to either show results or shut it down. McKinsey had previously run a $400k engagement that produced a 200-slide deck and zero behavioral change.

Diagnostic

The four-quadrant lens revealed the initiative was operating one developmental stage below the org's center of gravity. The UL (individual) and LR (system) quadrants had been addressed (training + tooling), but the LL (cultural) and UR (behavioral) quadrants were actively resisting. The org's incentive structure rewarded heroics over systematic improvement, so any initiative that didn't produce individual hero moments was unfundable at the team level.

Intervention

Redesigned the initiative's conditions rather than its messaging. Changed the incentive structure (LL quadrant) so that team-level adoption metrics fed into engineering-manager performance reviews, rather than relying on individual motivation. Removed the "transformation office" as a gating layer and redistributed its authority to existing engineering leads. Replaced the 200-slide deck with a 1-page "conditions for adoption" contract that each team signed.

Outcome

Adoption crossed 60% in 8 weeks (up from sub-15% over 18 months). The COO presented the engagement model at an industry conference (anonymized). The "transformation office" was dissolved. Total engagement cost: $18k. The previous McKinsey engagement: $400k.

Forge

AI-native startup

Engagement
MCP infrastructure audit
Size
$8k
Duration
2 weeks

Markers

  • Series A
  • $8M raised
  • 25 engineers
  • 2-week audit

Context

A Series A AI startup (25 engineers, $8M raised) had built their agent infrastructure on a tangle of LangChain chains, custom prompt templates, and ad-hoc function-calling. Latency was 4-8s per agent step. The CTO was evaluating whether to rebuild on a different framework or hire a dedicated ML infra engineer. Three vendors had pitched $80-150k "AI platform rebuild" engagements.

Diagnostic

The audit mapped the agent stack against the MCP (Model Context Protocol) specification. 7 of 12 agent capabilities were reimplementing MCP-standard tools (filesystem, search, code execution) with custom, less-tested code. The latency bottleneck was not the framework but redundant context-window rebuilding on every step. The "rebuild on framework X" pitch was solving the wrong problem.

Intervention

Replaced 7 custom tool implementations with MCP-standard servers (3 open-source, 4 from the existing codebase wrapped as MCP tools). Introduced a context-cache layer using a content-addressed store. Did NOT rebuild the framework. Total code change: ~2,000 lines removed, ~600 lines added.

Outcome

Latency dropped from 4-8s to 800ms-1.4s per agent step. Monthly infra cost dropped 40% (reduced context-token consumption). The CTO hired a generalist engineer (not a dedicated ML infra specialist) to maintain the MCP layer. Total engagement cost: $8k. The three vendor pitches: $80-150k each.

HoloOS

Healthtech

Engagement
Cultural diagnostic + PESTLE framework
Size
$22k
Duration
8 weeks

Markers

  • Growth-stage
  • 200 employees
  • 8-week engagement
  • FDA-blocked

Context

A growth-stage healthtech (200 employees, post-COB bridge round) had a clinical-engineering cultural split that was blocking their FDA submission. The clinical team (MDs, regulatory) and the engineering team (ML, platform) had stopped speaking to each other. Two previous "culture" consultants had run workshops that produced sticky notes and zero change. The CEO was considering replacing one of the two VPs.

Diagnostic

The four-quadrant lens mapped the cultural split as a UL/LL mismatch, not a personality conflict. The clinical team operated from a rule-following (LR-dominant) developmental stance; the engineering team operated from an achievement (UR-dominant) stance. Both were correct for their function. The conflict was structural, not personal. The PESTLE analysis showed the FDA submission was the political-economic (P+E) forcing function that made the conflict unresolvable without external intervention.

Intervention

Did NOT run a workshop. Instead, designed a "translation layer" role (a senior engineer with clinical background) whose only job was to translate between the two developmental stances. Wrote a 4-page "developmental contract" that explicitly named the two stances and gave each team permission to operate from their own stance while the translator handled the boundary. Did NOT replace either VP.

Outcome

FDA submission filed 5 weeks later (had been blocked for 4 months). Both VPs remained in role. The translation layer role became permanent. The CEO described it as "the first consulting engagement that didn't try to fix the people." Total engagement cost: $22k. The two previous culture consultants: $60k combined, zero outcome.

The first engagement is a 30-minute scoping call at no cost.

If your situation doesn't fit the diagnostic frame, you'll know in 30 minutes and it won't cost you anything. No slideware, no sales process, no follow-up sequence.

Request a Discovery Call

Cookie Preferences

We use cookies to enhance your experience. By continuing to visit this site you agree to our use of cookies. Learn more