Method
How the system works.
Most of the volume is reading: datasheets, manuals, filings, postings. A harness of several models does that reading in a controlled workflow and writes what it finds into fixed tables with the source kept. People set the criteria and review the cases the evidence cannot settle.
The flagging is still being tested. We publish no accuracy figures until an expert-labelled benchmark has run.
Criteria first, then documents, then a table.
For each product line, a field application engineer defines what a fit is: the ranges the part covers, the interfaces it serves, power and temperature limits, and the incumbent parts the line can realistically displace. The questions the system asks are derived from this and nothing else.
For each candidate product, datasheets, product pages, manuals, regulatory filings, teardowns and job postings are collected and kept. A fixed table per product is filled with values and the source text each rests on. A value the documents do not state is recorded as unknown, never filled in.
Several models run in a controlled workflow with fixed inputs and outputs per stage, so the same product and the same evidence can be re-run and compared. No stage writes free text into a record.
- Datasheets and product pages: what the product needs
- Regulatory filings and teardowns: what is inside, where public
- Job postings: what a team works with, and when
- Your own field knowledge: where competing parts already sit
Fixed questions, three routes.
A fixed set of questions is answered against the table, with fixed answer options and a record of the evidence each answer used. Technical fit and commercial timing are judged separately. Each product is routed as a supported match, a supported exclusion, or flagged for review because the evidence is thin or conflicting. An unknown is not a mismatch.
People review the flagged cases with the checks, the quotes and the open questions in front of them. Not every record is checked by an engineer; the ones the evidence cannot settle are.
Built and tested on public products against a stand-in FPGA profile. There are no client case studies yet. The FPGA screening page describes the system as it is, including what is not proven.
One product, read from its datasheet.
A 2 MP GigE machine vision camera, read from its two-page public datasheet in September 2026 against a stand-in FPGA profile. Three of the fields the screen needs most were not stated, so the product was flagged for review. A test on a public product, not a client result.
| Field | Extracted value | Datasheet text |
|---|---|---|
| Sensor | Sony IMX290, CMOS rolling shutter, 1/2.8" | Sensor model Sony IMX290 … CMOS, rolling shutter |
| Sensor-to-processor interface | not stated | |
| Resolution and frame rate | 1920 × 1080 at 56 fps | Max. frame rate 56 fps @1920 × 1080 Mono 8 |
| Host interface | Gigabit Ethernet, GigE Vision 2.0, GenICam | Data interface Gigabit Ethernet … GigE Vision V2.0, GenICam |
| External memory | not stated | |
| Total power | 2.1 W typical at 12 V | Power consumption Typ. 2.1 W@12 VDC |
| Operating temperature | 0 to 50 °C | Working temperature: 0 °C to 50 °C |
| Incumbent FPGA or processor | not stated | datasheet names no FPGA, SoC or processor |
| Route | Review | three key fields not stated |
Rules that hold for every record.
These apply to every record the system produces.
- Observed facts, hypotheses and open questions are kept separate.
- Missing evidence is not evidence of incompatibility.
- Technical fit and commercial timing are separate assessments.
- A product announcement does not prove an open component-selection window.
- Hiring or repository activity does not by itself prove a purchasing programme.
- A source quote supports a value; it does not guarantee the extraction or the interpretation.
- Fixed answer options do not establish calibrated probabilities or accuracy.
- Several pages repeating one source are one source.
Read further.
- The FPGA screen, the worked system as built
- How each kind of signal is detected and what it can establish
- What design-in intelligence is
FPGA product-fit screening · How signals are detected · Design-in intelligence · The service
Ask what the evidence can establish for your part