MFMODEL FOUNDRYStart a brief

Services

Three engagements. Published prices. No discovery theatre.

Every engagement below has a fixed fee, a defined end, and a list of what lands in your repository when it is over. The numbers are here so you can decide whether to call us before you spend an hour finding out.

01

Foundry Sprint

A fixed-fee technical discovery that answers one question honestly: can this actually be built, with your data, at a cost that makes sense?

What the engagement includes

  • A data reality audit against your actual records, not a sample we were handed
  • Baseline experiments on the real task, so the estimate rests on measurement
  • An architecture memo covering models, retrieval, infrastructure, and integration points
  • A costed build plan your team could hand to any competent vendor
  • A written go or no-go recommendation, including the case for not building it

What you keep

  • Data audit and feasibility findings
  • Baseline evaluation results
  • Reference architecture
  • Costed build plan and timeline
  • Recorded walkthrough with your team
Scope a Foundry Sprint
02

Build with us

We design, train, integrate, evaluate, and ship the production system, then stay on while it settles into real use.

What the engagement includes

  • A working system deployed into your cloud account, under your control
  • An evaluation suite that gates every deploy on quality, safety, and latency
  • Observability for cost, latency, and failure modes from the first release
  • Runbooks and handover documentation written for your engineers
  • A 30 day stabilization window after launch, included

What you keep

  • Production system and source code, owned by you
  • Evaluation harness and baseline scores
  • Monitoring dashboards and alerts
  • Operations runbook
  • Engineering handover sessions
Scope a Build with us
03

Enable your team

Your engineers keep the work. We bring the patterns, the evaluation discipline, and the review that stops the expensive mistakes.

What the engagement includes

  • Working reference implementations in your own repositories
  • An evaluation harness your team can extend without us
  • Architecture review of what you have already built
  • Live working sessions rather than recorded slideware
  • A written capability plan for the next two quarters

What you keep

  • Reference implementations and notebooks
  • Evaluation harness and documentation
  • Architecture review findings
  • Working session recordings
  • Capability roadmap
Scope a Enable your team

After launch

Operations retainer

$2,500 to $8,000 per month

A model that shipped is not a model that is finished. Data drifts, costs move, and edge cases arrive in the shape of angry users.

  • Monitoring review and drift detection
  • Evaluation refresh as your data changes
  • Model and prompt updates with before and after scores
  • Inference cost tuning
  • A named engineer who already knows your system

Honest scope

When not to hire us.

Every one of these has walked through the door. Saying so here saves us both a call, and it is a better use of your budget than finding out in month three.

  • You need a chatbot on your marketing site. That is a weekend of work and you should not pay a studio for it.
  • You have no data and no way to collect any. Models learn from evidence, and enthusiasm is not evidence.
  • The goal is an announcement rather than an outcome. We are the wrong partner for a press release.
  • You need the whole thing live in two weeks. Evaluation takes the time it takes, and skipping it is how these systems fail in public.
  • Nobody internally owns the workflow. Without a decision maker who wants this to work, it will not survive contact with the org.

Straight answers

Everything people ask before signing.

If your question is not here, ask it in the brief and we will answer it in writing.

What does an engagement actually cost?

A Foundry Sprint is $9,500 to $15,000 fixed. A production build runs $40,000 to $150,000 depending on integration surface and evaluation burden. Ongoing operations retainers are $2,500 to $8,000 per month. We publish these because a discovery call should be about your problem, not about extracting your budget.

Do you train your own models or use existing ones?

Whichever the measurements support. Most production value comes from retrieval, structured outputs, and careful evaluation around a strong existing model. Fine tuning earns its place when the task is narrow, repetitive, and well represented in your data. We show you the comparison rather than asking you to trust the choice.

Where does our data go?

Into your cloud account by default. We build in infrastructure you own, so your records stay under your contracts, your retention policy, and your auditors. If a hosted model is part of the design, we tell you exactly what leaves your boundary and what the provider's retention terms are before you approve it.

Who owns what you build?

You do. Source code, trained weights, evaluation suites, and documentation transfer to you, with no license back to us and no dependency on our hosting. If you fire us the week after launch, the system keeps running.

What happens if the model does not work well enough?

That is precisely what the sprint is for, and it is why we run baselines before quoting a build. If the evidence says the accuracy will not clear your bar, we recommend against building and you have spent a sprint fee rather than a build budget. If a build is already underway and a component underperforms, the evaluation gates catch it before your users do.

How involved does our team need to be?

Less than you would expect and more than you would like. Realistically: a decision maker for about an hour a week, and someone who genuinely understands the data and the workflow for a few hours a week during discovery and integration. Projects fail without that second person.

Do you work remotely?

Yes, and we work on site when the problem is physical. Inspection, line vision, and anything involving equipment usually needs someone standing in front of it at least once. We are based in Green, Ohio.

How quickly can you start?

We hold one build and one sprint at a time, deliberately, because attention is what makes these systems work. That means there is sometimes a wait. Ask and we will tell you the real date rather than the convenient one.

Next step

Tell us what is breaking down.

A brief takes about five minutes to write. We reply with an honest read on whether this is a sprint, a build, or something you should not pay anyone to do.