How datin works

See the same connected history from any sample, workflow or equipment.

Any way in, the same record Six ways in sit evenly around one record, and each points into it: Workflow, Sample, Equipment, People, Projects and Reports. The record in the middle is a small linked graph on a pale disc. In motion, the record appears, Workflow, Sample and Equipment arrive, and People, Projects and Reports join them. Then each way in lights up in turn, and the same record lights up from it. Workflow Projects Equipment Reports Sample People

datin in three parts

Your people and datin's agent work on the same record.

  • The context infrastructure

    One graph links results, samples, equipment and methods from every lab, so AI works across all your R&D.

  • Lab Workspace

    This LIMS and ELN module gives your scientists and engineers superpowers. Flexible enough for any research. Teams record and read data their way.

    See it in a test lab →

  • An autonomous scientist

    datin's agent works beside the people who made your company great. It reads your whole record, cites its sources and suggests what to test next.

Two ways into one record At the top a scientist at a tablet, Lab Workspace, has added a step to the record below: a teal line runs down from the tablet to a new node in a small teal graph of samples, equipment, methods and results. At the bottom datin's agent, the coral arrow in an app window, reads the same record: its coral line runs up to the new node, which with two older nodes is lit coral, and a small coral answer card sits on the scientist's tablet, its two source dots tied by threads to the two nodes it used. Lab Workspace Record Agent

What the right record needs

Where datin sits Formulation, materials testing, rig tests and automation all record into one record, datin, which holds samples, equipment, methods and results. The lab's existing systems, such as LIMS, ELN, ERP and files, stay beside it, linked with datin both ways. datin's agent reads the record in a small app window. LIMS, ELN, ERP, files datin's agent datin Samples Equipment Methods Results Formulation Materials testing Rig tests Automation
  • Record at the source

    Captured when it happens, not rebuilt for reports.

  • Readable by people and agents

    Technicians understand the structure. An agent queries it without guessing.

  • Grow one workflow at a time

    Start with the test that hurts most. Add the next without rebuilding.

  • Keep the history

    Who did what, when, and under which conditions.

Knowledge graphs, built for the lab

Web search, drug discovery and fraud detection already run on knowledge graphs.They connect what is known so questions follow the links.datin brings that to your lab.

  • Context for AI

    The memory your agents and robots read. Every answer has a source.

  • Every link has a meaning

    Results link to their sample, equipment and method so you can follow them.

  • Room to grow

    Add new test types to the graph without database changes.

  • Open standards
  • An ontology built with domain experts
  • Checked as data is written

How your data gets in

Start from what you already have.

  • Recorded at the bench, by people and robots
  • Instrument exports
  • Simulation runs, with inputs and code
  • Spreadsheets
  • LIMS, ELN and ERP exports
  • Preparation and characterization images
  • Datasets from past projects

Before

Two tables, one sample? Two lab tables name the same sample with slightly different keys. A synthesis log lists catalyst, temperature and sample ID, with S-0042 and S-0043. A table of GC results lists sample and yield, with s42 and s43. A thin line joins S-0042 and s42 and asks: same sample? Synthesis log Catalyst T Sample ID Pd/C 80 °C S-0042 Pt/C 80 °C S-0043 GC results sample yield s42 91 % s43 87 % same sample?

In datin

One sample in a graph One sample, S-0042, as a node in a graph, linked to its synthesis step with Pd/C at 80 °C, its GC result of 91 %, its instrument, the GC, and its method. S-0042 Synthesis: Pd/C, 80 °C GC result: 91 % Method Instrument: GC

In the product

  • Trace any result

    Every step is its own record: time, place, who or what ran it, inputs and outputs. Follow any result back to its specimen, machine and method.

    One result, traced back One result, a measured curve, with a coral thread from it back to what produced it: the specimen, the machine and the method. Result Specimen Machine Method
  • Analyze on the record

    Run Python on the record in the built-in coding environment. Each analysis becomes its own recorded step.

  • Lookups while you record

    Search more than 124 million PubChem compounds and the TIB Terminology Service inside the hosted app.

  • Tables across tests

    Results from different tests in one table, with their conditions as columns.

See it on your own data.

  1. 15-minute call
  2. Scoping
  3. One lab first, fixed fee

Deployed on-premises or in the cloud. Your data stays yours, and confidential.

Your contact

Dr. Nick Garabedian

Dr. Nick Garabedian

Founder & CEO, datin GmbH

Connect on LinkedInBook 15 min