Godela

What do they actually do

Godela is building an AI physics engine that gives engineers fast, simulation‑quality answers from their own data. Today it’s in early access with private demos and pilots; visitors can request access on the site, and public materials describe demos rather than a broadly available self‑serve product godela.ai YC company page American Bazaar.

In pilots, users upload CAD files, lab/test data, or past simulation results, then ask plain‑language “what‑if” questions. The system returns predicted outcomes and explainable relationships so teams can explore designs and root causes without running a full solver each time godela.ai YC launch.

Practically, they are not yet replacing entire simulation stacks in production. The current focus is proving accuracy and utility on real datasets with enterprise engineering teams, and building the integrations and infrastructure needed for broader rollout godela.ai American Bazaar.

Who are their target customer(s)

• Enterprise R&D engineers (automotive, aerospace, industrial): They run FEA/CFD/thermal workloads and wait days for solver runs and scarce compute/licenses, which slows design cycles and decisions. Godela targets faster, solver‑quality answers to cut these delays godela.ai American Bazaar.
• CAE / simulation teams managing large model batches: They face backlogs, high per‑run costs, and can’t explore many scenarios quickly. Godela’s pilots aim to speed scenario exploration and reduce dependence on full solver runs godela.ai YC launch.
• Mechanical and product design leads: They need quick, trustworthy tradeoff answers but currently wait for long simulations or build prototypes. Godela positions explainable, rapid outputs to help iterate without every physical prototype or full solver run godela.ai.
• Materials and lab R&D teams: It’s hard to fuse experimental data with simulation predictions, leading to many expensive tests. Godela’s workflow ingests lab/test data to produce physics‑aware predictions that can cut experiments godela.ai.
• Small hardware startups / lean engineering teams: Limited compute budgets and small teams make long runtimes and complex toolchains a bottleneck. Early access pilots promise faster iteration with lower overhead YC company page godela.ai.

How would they acquire their first 10, 50, and 100 customers

• First 10: Leverage founders’ network, YC intros, and targeted outreach to engineering leaders at a handful of automotive/aerospace/industrial firms to secure paid pilots with clear success criteria, delivered on customer data and converted to multi‑month trials.
• First 50: Package the first wins into repeatable “pilot kits” (data intake, eval checklist, timeline) and sell via SDR outbound to CAE teams plus targeted industry events; publish 2–3 case studies and enable a small channel of simulation consultancies to resell/run pilots.
• First 100: Formalize partnerships with CAD/PLM vendors, cloud providers, and SIs for co‑sell/OEM; launch a self‑serve tier for smaller teams with templates/vertical packs, and staff a lean growth team to convert inbound and drive multi‑project expansion.

What is the rough total addressable market

Top-down context:

Near‑term TAM aligns with CAE/solver software (~$8.9B in 2024) Fortune Business Insights. If the product generalizes to broader simulation/digital twin, the category is ~$23–24B (2024) Grand View Research, with longer‑term upside touching HPC spend (~$55–60B, 2024) Grand View Research.

Bottom-up calculation:

Start with CAE‑intensive enterprises in target verticals (auto, aero, industrial) that run frequent what‑if studies; estimate adoption as (number of suitable teams) × (pilot‑to‑production ACV), growing accounts as models validate and integrate into CAD/PLM workflows.

Assumptions:

• Enterprises will adopt AI surrogates for parts of their solver workflows once accuracy is validated on their data.
• Integrations into existing CAD/PLM and data pipelines are sufficient to fit current processes and compliance.
• Procurement will initially fund pilots/POVs before expanding to multi‑project licenses.

Who are some of their notable competitors

• Ansys: Incumbent CAE suite provider; rolling out SimAI and other AI‑augmented workflows that may satisfy customers inside the Ansys ecosystem without adding a new vendor Ansys SimAI.
• NVIDIA (PhysicsNeMo / Omniverse): Provides toolkits and runtimes (PhysicsNeMo, Omniverse, Warp) for AI surrogates and digital twins; enterprises or vendors can build on NVIDIA stacks instead of buying a separate solution NVIDIA PhysicsNeMo.
• SimScale: Cloud‑native CAE platform shipping ready‑to‑use AI physics models (built with NVIDIA tools) for near‑real‑time feedback on common problems, competing for engineers who want fast, cloud simulations press.
• Altair: Established simulation vendor packaging AI‑powered reduced‑order models and physics‑ML (e.g., PhysicsAI, romAI) within its suite, offering a speedup path inside existing tooling Altair AI‑Powered Engineering.
• Siml.ai / DimensionLab: Startup focused on training/deploying AI‑based simulators and interactive digital twins; integrates NVIDIA PhysicsNeMo and targets the same fast learned‑simulator use case NVIDIA feature.