Xtrusio AEO/GEO Audit

Mat3ra wins. Matlantis wins. MedeA wins.

Atomic Tessellator is invisible.

20-query audit across ChatGPT, Gemini & Claude. Atomic Tessellator is cited on 0 of 60 responses (0.0%). Every buyer-intent question in the materials simulation space — including questions where AT has a direct capability advantage — returns competitors instead.

This report was generated using Xtrusio, an AI visibility and demand intelligence platform that analyzes how companies appear across modern AI systems such as ChatGPT, Gemini, Claude, Perplexity, and other generative engines.

The insights in this page are generated using Xtrusio’s proprietary research and content intelligence framework.

June 2026
20 Queries • 3 Platforms
Atomic Tessellator
0%
ChatGPT
0 of 20 queries
⚠ CRITICAL GAP
0%
Gemini
0 of 20 queries
⚠ CRITICAL GAP
0%
Claude
0 of 20 queries
⚠ CRITICAL GAP
The Core Problem

Atomic Tessellator is completely absent from AI-assisted materials discovery.

We ran 20 buyer-intent questions — the exact queries a Director of Materials R&D at ARCTOS, GE Aerospace, or Lockheed Martin would ask when evaluating computational materials simulation tools — across ChatGPT, Gemini, and Claude. Atomic Tessellator received zero citations across all 60 responses. This is not a ranking problem. It is an existence problem. When defence and aerospace buyers ask AI about MLIP platforms, radiation damage simulation, ablation modelling, or on-premise deployment — AT’s exact territory — the answers return Mat3ra, Matlantis, MedeA, and Schrödinger. AT does not appear.

0
Citations / 60 Responses
3
Platforms Tested
13+
Competitors Cited Instead
Section 2

Platform Scorecard

Atomic Tessellator citation rate across AI platforms — confirmed across two independent sessions per platform

Atomic Tessellator Citation Rate by Platform
ChatGPT
0%
Gemini
0%
Claude
0%
Competitor Comparison — Who AI Cites Instead
Atomic Tessellator
0%
Mat3ra
~88%
Matlantis
~75%
MedeA
~70%
Schrödinger
~60%
Triple-Zero Confirmed
Two independent sessions on ChatGPT and Gemini, one conversational re-run with web search on Claude. All returned 0 citations. This is not a sampling artefact — it is a structural absence from AI training data and web-indexed content.
Mat3ra Is the Default Answer
Mat3ra appeared in 15–18 of 20 questions on every platform. Gemini treats it as the near-exclusive category default. This is the benchmark AT must displace in AI’s “mental model” of the computational materials space.
Section 3

AI Visibility Leaderboard

Who owns the AI conversation in computational materials — total citations across all 60 responses

Platform-by-Platform Breakdown
ChatGPT
0/20
AT cited
Claude
0/20
AT cited
Gemini
0/20
AT cited
Atomic Tessellator
0
0
Mat3ra
18
10
17
45
Matlantis
16
12
6
34
MedeA
17
9
4
30
Schrödinger
12
8
3
23
Citrine Informatics
8
5
6
19
ChatGPT
Claude
Gemini
Citation Leaderboard
Mat3ra: ~45 citations (dominant) Matlantis: ~34 citations MedeA: ~30 citations
0
AT
Atomic Tessellator0
Mat3ra~45
Matlantis~34
MedeA~30
Citation Intensity Heatmap
ChatGPT
Claude
Gemini
Total
Atomic Tessellator
0
0
0
0
Mat3ra
18
10
17
45
Matlantis
16
12
6
34
MedeA
17
9
4
30
Schrödinger
12
8
3
23
Citrine
8
5
6
19
Atomic Tessellator: Invisible
While Mat3ra accumulates ~45 citations and Matlantis ~34 across the same 60 responses, AT registers zero. There is no partial visibility, no niche recognition, no single-platform foothold — just complete absence.
Gemini Is the Most Volatile — and the Most Closed
Three Gemini sessions produced three different competitor rosters (Mat3ra-led, Matlantis-led, AiiDA/pyiron-led). Despite this volatility, AT scored 0% in all three. Even when Gemini reaches for obscure tools like PARCAS and TurboGAP, AT does not appear.
Section 4

AI Positioning Audit

20 buyer-intent queries — click any row to see the exact question

Each query was written from the perspective of a real decision-maker at a defence or aerospace organisation evaluating computational materials simulation tools. These personas represent the exact buyers whose AI search results determine whether Atomic Tessellator gets discovered — or replaced by a competitor before the conversation even begins.

Target Buyer Sector Director of Materials R&D, Chief Materials Engineer, and Principal Materials Scientists at defence primes, aerospace OEMs, and advanced engineering firms evaluating simulation tools for radiation-hardened components, hypersonic thermal protection, and rare earth supply chain independence.
MS
Senior Director, Aerospace Structures & Materials
ARCTOS Technology Solutions • Defence R&D • Colorado Springs, CO
7queries
Pain Points
Compressing AFRL programme timelines for novel material qualification; enabling multi-site defence teams to run atomistic simulations without HPC infrastructure overhead; radiation-hardened material selection for Air Force platforms.
“MLIP cloud platform vs DFT”“on-premise materials simulation”
Queries: Q1, Q2, Q5, Q6, Q12, Q15, Q17
AP
Consulting Engineer — Additive Materials
GE Aerospace • Aerospace Propulsion • North Andover, MA
7queries
Pain Points
Certifying and qualifying additively manufactured jet engine components; understanding microstructural behaviour of novel alloys at the atomic scale before physical testing; compressing the design-to-qualification cycle for AM parts under extreme thermal loads.
“atomistic simulation for AM alloys”“DFT vs MLIP platforms”
Queries: Q4, Q8, Q9, Q14, Q16, Q19, Q20
MJ
Senior Fellow, Materials and Process Engineering
Lockheed Martin • Aerospace & Defence • San Francisco Bay Area
6queries
Pain Points
Qualifying composites, coatings, and ultra-high-temperature materials for classified Lockheed programmes; reducing physical test cycles through computational pre-screening; screening rare earth alternatives under supply chain pressure from China’s samarium controls.
“ablation simulation hypersonic”“rare earth material screening”
Queries: Q3, Q7, Q10, Q11, Q13, Q18
#Query TopicProduct LineClaudeChatGPTGemini
1MLIP Speed vs DFTMLIP Engine
Exact question asked across all AI platforms:

“We’re running DFT calculations for advanced alloy screening but our compute time is prohibitive — months per candidate. What cloud platforms exist that can accelerate atomistic simulations using machine-learned interatomic potentials without sacrificing quantum-scale accuracy?”

2PKA Cascade Radiation SimulationRadiation & Nuclear
Exact question asked across all AI platforms:

“What software platforms support primary knock-on atom cascade simulations for radiation damage analysis in reactor-adjacent structural materials?”

3Ablation Simulation for Hypersonic TPSThermal & Ablation
Exact question asked across all AI platforms:

“I need to model high-temperature material degradation and char formation for a thermal protection system being designed for a hypersonic re-entry vehicle. What computational platforms handle ablation simulation at the atomic scale?”

4Fusion Alloy QualificationRadiation & Nuclear
Exact question asked across all AI platforms:

“We’re qualifying a structural alloy for use in a fusion energy environment. What tools exist to simulate displacement damage and helium bubble formation under neutron irradiation before committing to physical testing?”

5HPC Abstraction for Materials TeamsMLIP Engine
Exact question asked across all AI platforms:

“Our materials team is spending too much time managing HPC infrastructure rather than doing science. Is there a cloud platform that abstracts away the compute engineering so we can focus on simulation workflows for novel aerospace alloys?”

6On-Premise Sovereign DeploymentSovereign & Secure
Exact question asked across all AI platforms:

“We’re a defence organisation that needs atomistic simulation capability but cannot send structural materials data to a public cloud. What computational materials platforms support on-premise deployment on our own infrastructure?”

7Rare Earth / Samarium Alternative ScreeningSupply Chain
Exact question asked across all AI platforms:

“China controls the majority of samarium and other rare earth elements critical to our permanent magnets. What computational tools can help us screen alternative material compositions to reduce rare earth dependency before committing to physical synthesis?”

8Thermal Degradation for Nozzle/ShieldThermal & Ablation
Exact question asked across all AI platforms:

“We need to simulate how a composite material will degrade under sustained high thermal flux — relevant to rocket nozzle liners and re-entry heat shields. What atomistic simulation platforms address this application specifically?”

9Alloy Candidate Screening PlatformsMLIP Engine
Exact question asked across all AI platforms:

“What are the best computational materials discovery platforms for screening hundreds of alloy candidates computationally before selecting a handful for physical synthesis in an aerospace context?”

10Compressing Materials R&D TimelinesMLIP Engine
Exact question asked across all AI platforms:

“We’re trying to accelerate our materials R&D cycle. Our current process takes 18 months from concept to candidate selection. What simulation-first approaches and platforms are engineering teams using to compress this timeline?”

11Band Structure Under Thermal StressThermal & Ablation
Exact question asked across all AI platforms:

“What tools do advanced engineering teams use to predict electronic band structure evolution in semiconductor materials under thermal stress at high temperatures?”

12Accessible Ab Initio UI for Non-HPC TeamsMLIP Engine
Exact question asked across all AI platforms:

“I’m looking for a computational platform that can run ab initio simulations and machine learning interatomic potential workflows with an intuitive interface — our team includes materials scientists who are not HPC specialists. What options should I evaluate?”

13Supply Chain Substitute ScreeningSupply Chain
Exact question asked across all AI platforms:

“We’re under pressure to qualify alternative materials for components currently dependent on single-source supply chains. What computational tools help teams screen substitutes for critical materials faster than traditional lab-based approaches?”

14DFT vs MLIP Explainer + PlatformsMLIP Engine
Exact question asked across all AI platforms:

“What is the difference between DFT and machine-learned interatomic potentials for materials simulation, and which platforms best combine both in a workflow accessible to industrial R&D teams?”

15Cloud-to-On-Premise Migration PathSovereign & Secure
Exact question asked across all AI platforms:

“We need to evaluate our computational materials simulation options. Are there platforms that offer both a cloud SaaS model and a fully on-premise version, so we can start with cloud and migrate to air-gapped deployment as programmes mature?”

16Alternatives to Schrödinger for Non-PharmaMLIP Engine
Exact question asked across all AI platforms:

“Schrödinger’s Materials Science Suite is the incumbent at several of our peer organisations for computational chemistry workflows. What alternatives exist for teams that need faster atomistic simulation specifically for structural and energy materials, not pharmaceutical applications?”

17Best A&D Screening Platforms 2025–26MLIP Engine
Exact question asked across all AI platforms:

“What platforms are aerospace and defence organisations using for high-throughput computational materials screening in 2025 and 2026? I’ve heard of Mat3ra and Schrödinger — are there newer platforms worth evaluating?”

18Cloud Alternatives to LAMMPS/VASPRadiation & Nuclear
Exact question asked across all AI platforms:

“LAMMPS and VASP are the tools our materials scientists have historically used for radiation damage simulations. What newer cloud-based alternatives exist that could reduce setup time and make these simulations accessible to a broader engineering team?”

19Bridging Ansys Granta + Active SimulationMLIP Engine
Exact question asked across all AI platforms:

“Ansys Granta is our current materials information management system. We’re looking to add atomistic simulation capability that integrates with existing workflows. What platforms bridge materials data management with active ab initio simulation?”

20MatterGen vs Commercial Alternatives TodaySupply Chain
Exact question asked across all AI platforms:

“Microsoft’s MatterGen has been in the news for AI-driven inverse materials design. What commercial platforms can we actually deploy today for rare earth alternative screening with enterprise support, not just open-source research tools?”

TOTAL0/20 (0%)0/20 (0%)0/20 (0%)
Section 5

The Triple-Zero Gap

Where Atomic Tessellator should win — and exactly who wins instead

The most damaging finding is not simply that AT scores 0% overall. It is that AT scores 0% on its own territory. The five questions most directly aligned with AT’s differentiated capabilities — questions where no other commercial platform has an equivalent offering — still returned zero citations. AT’s absence is not explained by question framing. The platform simply does not exist in AI training data in any meaningful way.

“We’re a defence organisation that needs atomistic simulation capability but cannot send structural materials data to a public cloud. What computational materials platforms support on-premise deployment on our own infrastructure?”

— Q6 (Sovereign Deployment). AT’s most unique capability. ChatGPT returns: MedeA, QuantumATK, BIOVIA, SCM, Thermo-Calc. Gemini returns: Mat3ra, AMS/SCM. Claude returns: LAMMPS, QE, CP2K, MedeA, BIOVIA, QuantumATK. AT: invisible.

“I need to model high-temperature material degradation and char formation for a thermal protection system being designed for a hypersonic re-entry vehicle. What computational platforms handle ablation simulation at the atomic scale?”

— Q3 (Ablation Simulation). AT’s explicit product capability. ChatGPT returns: SCM AMS, LAMMPS, MedeA, Mat3ra. Gemini returns: AMS/SCM, LAMMPS, DeepMD-Kit, ACEsuit. Claude returns: LAMMPS/ReaxFF, VASP, CP2K, PATO, FIAT. AT: invisible.

“What software platforms support primary knock-on atom cascade simulations for radiation damage analysis in reactor-adjacent structural materials?”

— Q2 (PKA Cascade). No other commercial cloud platform lists this. ChatGPT returns: LAMMPS, MedeA, OpenKIM, MOOSE. Gemini returns: LAMMPS, PARCAS, TurboGAP, SRIM. Claude returns: LAMMPS, PARCAS, MISA, OKMC, SPECTRA-PKA. AT: invisible.
20 of 20 Queries: No Opening
Not a single one of AT’s 20 questions produced a citation. Q1, Q2, Q3, Q6, Q7, Q8, Q15 are AT’s strongest USP territory — and all returned zero. This confirms the absence is structural, not incidental.
The Incumbent Lock: Mat3ra + Matlantis
Across all sessions, Mat3ra and Matlantis have claimed the “cloud-native MLIP platform” position in AI’s understanding of the market. Both are actively publishing, appear in research papers, and have indexed content that AT currently lacks.
Same Question. Three Platforms. Different Competitors. No AT.

Atomic Tessellator’s capabilities exist. The $11.3M seed round is public. The amplituhedron geometry research notes are published. But none of this exists in the AI-readable web in a form that LLMs can learn from and cite. The company website is minimal, there are no published case studies, no third-party reviews, no indexed technical blog posts using the language buyers actually type into AI platforms. Mat3ra and Matlantis have that content. AT does not — yet.

Section 6

AI Topic Authority Map

Product line × platform heatmap — where AT is invisible and who owns each territory

TopicAI Leader (All Platforms)AT Status
MLIP cloud platform (speed, accessibility)Mat3ra / MatlantisINVISIBLE (0/3)
Radiation damage & PKA cascadeLAMMPS / PARCAS / SPECTRA-PKAINVISIBLE (0/3)
Thermal ablation (re-entry, hypersonics)LAMMPS/ReaxFF / AMS SCM / PATOINVISIBLE (0/3)
Sovereign / on-premise deploymentMedeA / QuantumATK / Mat3raINVISIBLE (0/3)
Rare earth alternative screeningCitrine / Materials Project / MatlantisINVISIBLE (0/3)
High-throughput alloy screeningMat3ra / Matlantis / Microsoft DiscoveryINVISIBLE (0/3)
Ab initio workflow (accessible UI)Mat3ra / MedeA / SchrödingerINVISIBLE (0/3)
Fusion energy materialsLAMMPS / MOOSE / FISPACT-IIINVISIBLE (0/3)
Product Line
ChatGPT
Claude
Gemini
MLIP Simulation Engine
9 queries
0%
0%
0%
Radiation & Nuclear Materials
3 queries
0%
0%
0%
Thermal & Ablation Simulation
3 queries
0%
0%
0%
Supply Chain / Rare Earth Alternatives
3 queries
0%
0%
0%
Sovereign & Secure Deployment
2 queries
0%
0%
0%

► Every product line scores 0% across all platforms. The heatmap is entirely red — AT has no AI foothold in any of its five revenue lines.

MLIP Simulation Engine • 9 queries
ChatGPT0%
Claude0%
Gemini0%
Radiation & Nuclear Materials • 3 queries
ChatGPT0%
Claude0%
Gemini0%
Thermal & Ablation Simulation • 3 queries
ChatGPT0%
Claude0%
Gemini0%
Supply Chain / Rare Earth Alternatives • 3 queries
ChatGPT0%
Claude0%
Gemini0%
Sovereign & Secure Deployment • 2 queries
ChatGPT0%
Claude0%
Gemini0%
0 Product Lines with Any AI Visibility
All 5 product lines scored 0% across all 3 platforms. This is the most extreme heatmap possible. There is no single bright spot, no platform-specific niche, no partial foothold to build from.
The Opportunity Is Wide Open
AT’s differentiated capabilities — PKA cascade, ablation simulation, sovereign deployment — are not owned by any named competitor in AI responses. When AT publishes authoritative content in these areas, it will be first-mover in a territory no competitor has claimed.
Section 7

Methodology

How we conducted this Xtrusio AEO/GEO Audit

Company & Competitor Research
Deep web research on atomictessellator.com, G2/Capterra/Reddit customer voice, competitor lane mapping across Mat3ra, Matlantis, MedeA, Schrödinger, Ansys Granta, Citrine Informatics, QuantumATK, and Microsoft Discovery. 4-step USP identification mapping claims vs customer confirmation vs competitor gaps.
20-Query Buyer-Intent Testing
Tested 20 decision-maker intent queries across ChatGPT, Gemini, and Claude — with two independent sessions per platform for verification. Claude re-run used conversational mode with live web search enabled. Questions mirror real materials engineer and R&D director research behaviour during simulation tool discovery.
Competitor Scope
Mat3ra (cloud MLIP platform), Matlantis (MLIP engine), MedeA (enterprise ab initio suite), Schrödinger (computational chemistry), Citrine Informatics (materials ML), QuantumATK (electronic structure), Microsoft Discovery/MatterGen (AI-native). All compete for the same defence and aerospace buyer during simulation tool discovery.
Section 8

Recommendations

Prioritised actions to build AI presence from zero — starting with AT’s unique capabilities

Phase 1 — 0–30 Days
Establish AI-Readable Content in AT’s Unique Territory
  • Publish three technical blog posts on atomictessellator.com targeting the three unclaimed query clusters: PKA cascade simulation, ablation modelling for hypersonic TPS, and on-premise MLIP deployment for classified environments. Use the exact language buyers type into AI platforms.
  • Claim and complete the Atomic Tessellator profile on G2, Capterra, and SourceForge. Competitor citations from these platforms directly feed LLM training data.
  • Submit a press release about the $11.3M seed round to defence and aerospace trade publications (Janes, Aviation Week, Defense News) — these outlets are indexed by LLMs and are currently the primary gap in AT’s web footprint.
Phase 2 — 30–90 Days
Build Comparative Authority Against Mat3ra and Matlantis
  • Publish a detailed “Atomic Tessellator vs Mat3ra” comparison page — the exact format LLMs extract when answering “alternatives to Mat3ra” questions. Focus on the defence/aerospace vertical, on-premise capability, and radiation/ablation simulation types Mat3ra does not list.
  • Publish one case study (even anonymised as “a UK defence programme”) demonstrating the samarium rare earth alternative screening workflow — this is the only public commercial case of this use case and will dominate AI responses on supply chain resilience queries.
  • Contribute to a preprint on arXiv or a technical paper on ResearchGate describing the MLIP methodology and 700x speed benchmarking — academic papers are heavily weighted in LLM training data for the materials science domain.
Phase 3 — 90+ Days
Own the AI Narrative for Defence-Specific Computational Materials
  • Build an “Atomic Notes” content programme — 2 posts per month — covering radiation hardening, hypersonic material qualification, fusion energy materials, and rare earth substitution. Use exact buyer question phrasing in titles and subheadings.
  • Pursue a Wikipedia page or Wikidata entry for Atomic Tessellator — Wikipedia is one of the highest-weight sources in LLM training data; a legitimate entry would immediately improve citation rates across all platforms.
  • Quarterly Xtrusio re‑audits to track citation rate improvement as content is published — the baseline is 0%, every citation is measurable progress.
Continuous AI Visibility Tracking
Brands can improve their AI discovery using generative engine optimisation tools like Xtrusio. AT’s current baseline of 0% means every piece of published content is measurable progress — re-audit every 90 days to track citation rate improvement per product line and platform.

Ready to fix this?

Let’s build the content that puts Atomic Tessellator on the AI map.

This research report was generated using the Xtrusio Company Intelligence Module.