Contact

Start with the microscopy workflow problem — not a generic software brief.

The best conversations begin with a concrete bottleneck: what the instrument acquires, what's failing today, and what outcome matters. NeuralSoftX returns with a feasibility read and, where it fits, a scoped proposal.

Primary contact

ivan.lobato@neuralsoftx.com

Email is the fastest path. Used for new project enquiries, collaboration discussions, and technical conversations about a specific microscopy workflow.

Open email app

If clicking the email link does nothing on your machine, use the copy button and paste the address into your preferred mail client.

A useful first message

What to include

  • Data type: SEM, STEM, TEM, CBED, 4D-STEM, or tomography
  • What is failing today: drift, non-rigid motion, low dose, noise, reconstruction quality, segmentation throughput, or operator bottleneck
  • Desired outcome: algorithm, deployable tool, workflow integration, or feasibility assessment
  • Practical constraints: on-prem requirement, dataset size, limited labels, or instrument-specific variation

How engagements usually work

A short, predictable first step.

1 — Intro call

A 20–30-minute call to confirm the workflow problem is in scope and identify the data characteristics that matter.

2 — Feasibility read

A written assessment of whether the problem is solvable, which machinery fits, and what level of effort the first milestone looks like.

3 — Scoped proposal

If the feasibility read is positive, a proposal with milestones, deliverables, and pricing. Engagements are structured around client milestones with clear written reporting at each step.

4 — Delivery

Algorithm, ONNX-packaged tool, workflow integration, or custom model — depending on what was agreed. The engagement ends when the workflow runs at the client's site.

Typical enquiry types

What tends to be in scope.

Registration & drift correction

Time-series SEM, STEM, or TEM data that changes too much during acquisition for standard rigid alignment.

Tomography preprocessing

Projection restoration, alignment, and compensation work before reconstruction quality degrades.

Restoration & segmentation

Low-dose image enhancement, particle or fibre analysis, and quantitative microscopy support workflows.

Simulation-backed development

Projects where experimental labels are scarce and MULTEM-generated synthetic data creates the training leverage.