Case Study 1 — Three routes, one COGS target

Context. A development-stage compound needed a GMP-compatible manufacturing route with a hard commercial constraint: cost of goods below €3,000/kg. The incumbent approach could not credibly reach it.

What we did. Rather than optimise a single route, we developed three independent synthetic strategies in parallel and costed each one. Each was built as a standalone dossier with full reaction schemes, step-by-step COGS breakdown, and an assessment of scale-up and supply risk.

The three shared a common late-stage strategy but diverged sharply in approach — an olefination-based route, a phosphonate-based route, and a transition-metal-coupling route — so the client could weigh cost against technical risk rather than being handed a single answer.

Outcome. All three routes came in substantially below the €3,000/kg target, the best at roughly a quarter of it. The client received not one recommendation but a decision framework: the cheapest route, the lowest-risk route, and the most IP-defensible route, each with its trade-offs made explicit.

Why it mattered. A single proposed route is a bet. Three costed routes are an option set — and options are what survive an investment committee.

Case Study 2 — Designing an asset that did not exist yet

Context. An open innovation challenge in autoimmune disease called for the rational design of a targeted degrader-antibody conjugate — combining targeted protein degradation with antibody-mediated delivery, a modality with very little precedent to draw on.

What we did. We worked the problem from both ends: the biology of target selection and degradation mechanism, and the chemistry of linker design, conjugation strategy, and payload stability. The design had to be scientifically defensible and plausibly manufacturable — a constraint most conceptual designs ignore.

Outcome. The submission was selected as the winning solution.

Why it mattered. This is the same capability applied in scouting: assessing whether a novel modality is genuinely tractable, or whether it is elegant on a slide and impossible in a reactor.

Case Study 3 — Building the monitoring layer

Context. Staying current across global pharma and biotech dealmaking is a volume problem before it is an analysis problem. Manual monitoring does not scale, and commercial platforms surface everything without weighting anything.

What we did. We built an automated intelligence pipeline that continuously ingests global pharma and biotech sources, with parallel streams for global activity and for specific regional markets. The system handles collection, deduplication and structuring; a human applies BD&L judgement to what surfaces.

Outcome. A running monitoring capability that produces qualified, prioritised intelligence rather than raw feed volume — now the engine behind our Innovation Watch service.

Why it mattered. Most scouting practices either drown in manual effort or outsource judgement to a database. We separated the two: machines for coverage, expertise for relevance.