Strata is Subterrix’s double-filtration AI engine that transforms a single query into a full field report. Your question first passes through Claude Haiku, which decomposes it and routes it simultaneously across six specialist lanes — covering numismatics, historical records, terrain, soil, and more. Five AI models process each lane in parallel. Claude Sonnet then fuses every output into one cited, coherent answer. The deeper mechanics of each layer reveal exactly why this architecture outperforms anything a solo detectorist could replicate manually.
Key Takeaways
- Strata uses a double-filtration AI architecture where Claude Haiku decomposes queries and routes them through six specialized research lanes simultaneously.
- Five AI models process each lane in parallel, covering vision, numismatics, historical records, land conditions, and terrain analysis.
- Claude Sonnet fuses all lane outputs into one cited, coherent answer rather than fragmented, general responses.
- Strata cross-references 1.8 million verified historical records alongside live soil, weather, and satellite data for field-ready insights.
- USGS LiDAR, Sentinel-2 imagery, and mineralization mapping help identify precise target areas with an 80% historical proximity score.
The Research Problem Every Detectorist Hits Before the First Dig
Before you drive to a site, you’ve already lost hours to a research problem that most detectorists never fully solve. Historical context isn’t optional — it’s the difference between a productive hunt and a wasted afternoon. You need land ownership records, colonial-era maps, troop movements, and settlement patterns all cross-referenced before you pull a single target.
Then there’s equipment calibration. Soil mineralization, drainage depth, and ground composition directly affect your detector’s performance, yet most hunters set their machines the same way regardless of conditions.
These two gaps — fragmented historical context and uninformed equipment calibration — define the pre-dig failure point. Strata was built specifically to close both.
It doesn’t simplify the research process. It executes it at a scale no individual detectorist realistically can.
How Strata’s Double-Filtration System Turns One Query Into a Full Report
When you submit a query to Strata, it doesn’t pass through a single AI model — it enters a double-filtration architecture that systematically breaks your question apart, routes it through specialist lanes, and rebuilds it into a single cited answer.
Claude Haiku handles the first filtration stage, decomposing your query and dispatching it across six specialist lanes: Vision, Numismatic, Historical, Research, Land, and Field Conditions.
Five AI models run in parallel across those lanes, each processing a specific dimension of your question. That’s where algorithm optimization drives precision — each lane returns structured output rather than generalized noise.
Claude Sonnet then executes the second filtration stage, handling data fusion by reassembling every lane’s findings into one coherent, cited response. You get depth without doing the routing yourself.
The Six Specialist Lanes Behind Every Strata Answer
Each of the six specialist lanes Claude Haiku dispatches your query to serves a distinct analytical function, and understanding what each one does explains why Strata’s outputs carry the depth they do.
Every lane applies machine learning to its designated domain, then feeds results into a data fusion layer where Claude Sonnet assembles a single, cited answer.
- Vision & Numismatic — Identifies artifacts and coins using parallel AI agreement-based scoring
- Historical & Research — Cross-references 1.35 million verified records to surface factual, documented insights
- Land & Field Conditions — Analyzes terrain, LiDAR elevation data, and real-time soil and weather conditions
No lane operates in isolation. Each one contributes a specialized signal, and the fusion stage ensures nothing gets lost between disciplines.
What Strata Finds in the Ground Before You Ever Step on the Site
Most detectorists walk a site blind, relying on intuition and surface-level research. Strata changes that before you ever lace your boots.
Using USGS 3DEP LiDAR data, Strata processes elevation models to detect ground dips, compacted travel paths, and structural depressions invisible at surface level. It maps geological layers to flag mineral deposits that distort signals and calculates how those variables affect your specific detector’s performance.
Modified A* pathfinding algorithms reconstruct historical movement corridors through the terrain, pinpointing where human activity concentrated over centuries.
The system then scores each zone through a compound confidence model, applying an 80% historical proximity weight. You receive a ranked, evidence-backed target area — not a guess. You’re walking in with a data advantage most detectorists don’t know exists.
Live Weather, Soil, and Satellite Data Strata Adds to Every Report
Knowing what’s beneath the surface only gets you so far — the conditions above it determine whether that data translates into productive detecting. Strata closes that gap through real-time data integration, pulling live weather, soil drainage metrics, and Copernicus Sentinel-2 remote sensing imagery into every report. You’re not guessing at field conditions — you’re reading them through verified, current data streams.
What this layer adds to your session:
- HuntCast scores live detecting conditions using real-time weather and soil data specific to your location
- Soil profiling delivers drainage depth and type, then cross-references your registered detector for targeted recommendations
- Sentinel-2 satellite imagery flags surface anomalies that correlate with subsurface archaeological potential
Every variable gets measured before you move.
Run Strata on Your Own Sites
Strata is the engine behind the research, and it’s built to be run, not just read about. Treasure Valley Metal Detecting Club members get Subterrix Elite, Strata included, for $8.99 a month instead of the standard $15.99, with 20% of every membership coming back to the club to fund hunts, raffles, and giveaways.
Join Subterrix under TVMDC for $8.99/month
Disclosure: TVMDC earns a share of membership revenue when you join through this link, at no extra cost to you.
Frequently Asked Questions
How Many AI Models Does Strata Run Simultaneously Across Its Specialist Lanes?
Strata runs five AI models simultaneously across its six specialist lanes, giving you unmatched AI model diversity and built-in research validation as each model actively contributes to synthesizing accurate, cited, multi-source intelligence.
Can Strata Generate Personalized Detector Settings for My Specific Machine?
Yes, Strata’s FieldIQ tool handles machine customization for your specific detector, delivering AI personalization across 15 supported machines with step-by-step settings instructions tailored to your registered device and local soil conditions.
How Accurate Is Strata’s Artifact Identification During Backend Metal Composition Scans?
Nine out of ten times, Strata’s artifact detection accuracy holds during backend metal composition analysis. It’ll correctly identify what you’ve found, then pinpoint specific detection locations based on terrain data—giving you verified, actionable intelligence in the field.
How Many Historical Records Does Strata Validate Its Research Findings Against?
Strata validates your research findings against over 1.35 million historical records, ensuring historical accuracy through rigorous data validation. You’re cross-referencing a verified database that keeps every insight grounded in confirmed, factual information—never fabricated or embellished.
What Pathfinding Algorithm Does Strata Use to Reconstruct Historical Travel Routes?
Strata uses a modified A* pathfinding algorithm for route reconstruction, processing USGS elevation data through advanced graph algorithms to map historical travel paths with pinpoint precision — revealing terrain secrets that’ve been buried for centuries.
References
- https://www.youtube.com/watch?v=AKNhNCS9MYk
- https://www.youtube.com/watch?v=e-2H0Gnq6Jg
- https://www.youtube.com/watch?v=ZbMZTAVw1i0
- https://focusspeed.com/subterrix-addressing-misconceptions/
- https://www.youtube.com/watch?v=MeR8_LKCBBE
- https://www.youtube.com/watch?v=1r74Bin55Qw
- https://www.subterrix.com/about
- https://www.subterrix.com/g2
