Defined by use
Every term here is defined the way we use it, not the way a textbook would.
Where a measure belongs to someone else, as the Intelligence Index belongs to Artificial Analysis, we say whose it is.
A
- Agentic workflow
Multi-step agents that act inside your tools, with guardrails and a person signing off where it matters. We put them in the workflows that cost the most hours.
Used on Forward deployed engineering
- AI due diligence
An operator’s read on a target’s data and AI position: what is real, what it would take, where the value is, and where AI threatens the model.
Used on AI due diligence, Private equity
B
- Build-ready brief
The step between discovery and build: a scoped brief, sequenced by dollar impact and risk, with success criteria set before the first line of code. The build then runs in a one to six week window.
Used on Get started, How we engage
D
- Data monetization
Turning the data a business already creates into products customers pay for.
Used on Private equity, Industries
- Data product
The pipelines and data models underneath an AI system, built to be used, owned, and measured.
Used on Forward deployed engineering
- Discovery workshop
The step after intake with the business sponsor: where generative AI, agents, and data products create value for you, and where they don’t.
Used on Get started, How we engage
E
- Evaluation harness
Tests of model quality on your own cases, run on every change, so quality is measured rather than assumed. It also decides between the models on a shortlist, and we re-run it when a better model ships.
F
- Forward deployed engineer
A senior engineer who works inside your business to scope, build, and ship a production AI system, and stays accountable until it is used. The people who scope the work are the people who build it.
Used on Forward deployed engineering
- Frontier model
A leading model from any lab, closed or open-weight, judged on independent benchmarks of intelligence, speed, and price. The frontier moves every few weeks, so we build with whichever frontier model fits the job and choose on evidence. On our charts, the frontier is the line through the most intelligence available at each price, or at each speed.
Used on Frontier models
H
- Hold period
The time a sponsor owns a portfolio company, and the span our AI and data roadmap is sequenced for.
Used on Private equity
I
- Intelligence Index Artificial Analysis
The Artificial Analysis Intelligence Index combines a set of independent evaluations, covering reasoning, knowledge, coding, and agentic work, into one score. Where a model runs at several reasoning settings, our frontier models page shows its best-scoring setting.
Source: Artificial Analysis (artificialanalysis.ai)
Used on Frontier models, Model explorer
M
- MCP workflow Model Context Protocol
Model Context Protocol servers that connect your systems once, for every model you use, so a switch of model is a configuration change, not a rebuild.
Used on Forward deployed engineering
P
- Playbook
The method behind a win, kept so it can be repeated. Every win becomes a reusable playbook, so the second deployment costs a fraction of the first; across a portfolio, a win in one company becomes the playbook for the next.
Used on How we engage, Private equity
R
- RAG pipeline Retrieval-augmented generation
Retrieval over your documents and data, with sources cited, so every answer can be checked.
Used on Forward deployed engineering
- Reference implementation
Each build documented as a reusable pattern, so the second deployment costs a fraction of the first. You get it with the working system and its playbook.
Used on Forward deployed engineering
S
- Success criteria
The measures a system is judged on, set before the first line of code: revenue, margin, hours reclaimed, adoption. Results are measured, not asserted, and we stay through rollout so the system is used.
Used on How we engage, Get started
T
- Technical standard
The choices we set on every build: orchestration, evaluation, and vector store, and a clear call on what stays in house and what goes to a partner. Across a portfolio, one standard holds for every company.
V
- Value creation plan
The plan a sponsor sets for a portfolio company’s hold period. Our part is the AI and data roadmap inside it, sequenced by dollar impact and risk, and tied to revenue, margin, and hours reclaimed.
Used on Private equity
Next
Tell us what you want to build. We will set the success criteria before the first line of code.