What happens after you open a conversation.
Every Lualdi engagement follows the same lifecycle — from first contact through deployed runtime and ongoing support. No surprises, no opaque pricing, no autonomous writes. The firm responds within two business days, produces a written scoping memo before any commitment, deploys inside the client's infrastructure, and hands off operational ownership when the runtime is stable. Below: the full path, the three engagement shapes, and what the firm asks of the operator.
The engagement is the same regardless of size or sector. What changes is the depth of each stage — a single-runtime deployment compresses; a custom build extends; an applied research note skips deployment and goes straight to a written report with a replayable run log.
You open a conversation — via the engagement form, an introduction, or a direct referral. Tell the firm who you are, what you operate, and what you're trying to decide better. The firm responds within two business days with triage and a discovery-call calendar invite. If the firm is not the right fit, you'll hear that directly with a referral elsewhere when possible.
A 45-minute video call with the founder and the relevant practice lead. You describe the current state — the decision points, the data you already have, the frictions in your existing process, the time pressure. The firm names which runtime likely fits (SIGMA · MITHRIL · ORIGINS · CARRIER) or whether a custom build is warranted, and what a proof-of-value would look like. No pricing yet, no commitment yet — pricing comes after scoping, when the work is well-defined enough to size.
The firm produces a written scoping memo. It includes: the problem statement in the firm's terminology, the proposed approach, what would be measured (and how), the runtime configuration, the data pipeline architecture, what you need to provide (access, sample data, point of contact), timelines, fee structure, and milestones. The scoping memo becomes the basis of the engagement letter. Signed → work begins.
The runtime is deployed inside your infrastructure (or, more commonly, into a sandboxed mirror first, for derisking). The firm declares your operational ontology in the runtime, connects the data pipelines, runs calibration cycles, and starts shadow runs alongside your existing process. Your team is trained on the daily/weekly digest format — what each output means, what to sign, what to defer. Shadow runs continue until performance is documented against your existing baseline.
First operator-signed decisions go to production. Performance baseline established. The replay log is archived with the engagement record. The firm hands off operational ownership to your team — the runtime is yours, running inside your infrastructure. The firm steps back from active operation, but remains accessible for support.
Quarterly model refits, monthly office hours with your operator team, drift monitoring on the model's performance, new industry packs added as your scope grows. Annual review of the engagement — what worked, what didn't, what to extend. The firm remains accessible but does not run the operation. The system is the operator's; the firm is the operator's partner in maintaining it.
Most engagements take one of three shapes. The lifecycle above applies to all three; what differs is the depth and duration of each stage.
You want one of the firm's flagship runtimes deployed and tuned to your data — SIGMA, MITHRIL, ORIGINS, or CARRIER. The firm declares your ontology, connects your data pipelines, and configures the runtime for your specific operations. Typical timeline: 6–10 weeks to live operation. Most common engagement shape.
You have a decision problem the flagship runtimes don't cleanly fit. The firm builds a custom decision system from the same core IP, configured around your domain. Examples: a new SIGMA configuration for a sector the firm hasn't packed; a MITHRIL ontology pack for an industry not in the standard six; a procurement runtime for a non-coffee commodity. Typical timeline: 12–24 weeks.
You have a question the firm should model end-to-end and answer in writing — typically a strategic, capital-allocation, or sector-positioning question. The firm produces a research report with a replayable run log, sized to whatever depth the question warrants. No deployment, no runtime; output is a written, sourced, replicable answer. Typical timeline: 4–12 weeks.
The firm cannot do its work in a vacuum. A successful engagement depends on the operator providing a handful of things — none individually heavy, but each essential. The scoping memo will name these specifically for your engagement; the list below is the universal baseline.
One person on your side who owns the engagement, has the authority to make scoping decisions, and will be the operator signing the runtime's outputs. Without a named operator with sign-off authority, the approval-gated model breaks down. The firm will not proceed past discovery without this.
Read access to the systems of record the runtime needs (ERP, OMS, TMS, market data feeds, sensor streams, whatever is relevant). The firm never moves your data outside your perimeter — the runtime is deployed inside your infrastructure — but it needs to see your data to ground its outputs. NDA and DPA are signed at engagement letter, before any data access.
What does the current process look like? What decisions are being made, by whom, on what cadence, against what metrics? The firm needs to understand the baseline in order to show what the runtime improves. If the baseline isn't documented, the firm will help you document it as part of discovery.
Every Lualdi runtime is approval-gated. Your team must be willing to operate inside that model — reviewing outputs, signing decisions, retaining the audit trail. This is not optional. The firm does not build runtimes that write autonomously, and will decline engagements where the operator wants full automation without human approval.
Engagement by introduction or referral · Response within two business days