What Adam users actually want
Mimir analyzed 15 public sources — app reviews, Reddit threads, forum posts — and surfaced 14 patterns with 7 actionable recommendations.
Top recommendation
AI-generated, ranked by impact and evidence strength
Build production validation scoring into the export flow with a confidence badge system for manufacturing-ready outputs
Rationale
Users must manually verify all AI-generated outputs before manufacturing (per Terms of Service), but the product provides no built-in validation mechanism. Engineers and 3D printing users are blocked from confidently moving designs into production, creating friction at the most critical conversion point. The team already has production scoring criteria defined (parts compile, constraints solve, references stay stable, manifold geometry) but hasn't surfaced this to users.
Without this, users who need manufacturing confidence — the core engineering and maker segments driving retention — are forced to manually inspect every output or abandon the tool for production work. This directly undermines the 10x speed value proposition if users spend validation time equivalent to manual CAD work.
Implement a three-tier badge system (Production Ready / Needs Review / Experimental) displayed at export, running the existing production scoring criteria in the background. Show which criteria passed/failed with one-click remediation suggestions for common issues (e.g., 'Constraint conflicts detected in 3 features — click to review'). This transforms a trust barrier into a feature that differentiates Adam from legacy CAD tools that provide no AI-assisted validation.
More recommendations
6 additional recommendations generated from the same analysis
The product serves three distinct segments (engineering teams, creators, 3D printing hobbyists) with different needs, but presents a single undifferentiated interface and feature set. Users landing from segment-specific pages (Engineering, Creators, 3D Printing) encounter the same generic editor, forcing them to discover relevant capabilities through trial and error. This creates unnecessary activation friction and likely contributes to early churn across all segments.
The product positions itself as a CAD copilot integrated into existing workflows (Onshape extension live, ambition to support other CAD tools), but there is no visibility into which integrations users need most or which would unlock enterprise deals. The Enterprise tier includes custom integrations and finetuned models, but without systematic demand signal collection, the team is building integration roadmap in the dark.
The service operates in beta with explicit disclaimers about instability, data loss risk, no uptime guarantee, and $100 liability cap. These terms are reasonable for early-stage software but create trust barriers for exactly the users most valuable to retention: engineering teams and professional creators who want to integrate Adam into production workflows. Users who experience value are blocked from deeper commitment by reliability uncertainty.
The core value proposition is replacing click-based CAD with natural language prompts, but users arriving from traditional CAD backgrounds do not know how to translate their intent into effective prompts. This creates activation friction: users who would benefit most from the AI-native interface (engineers accustomed to legacy CAD) are least prepared to use it effectively. The product expects users to invent their own prompt syntax, but this is a learned skill.
AI-generated CAD outputs are non-deterministic and users frequently need to iterate through multiple attempts to reach a usable design. The product promotes instant variations and prompt-based refinement, but if a generation fails or produces an unusable result, users lose context and must start over. This creates perceived risk in using AI for multi-step designs and discourages the exploratory workflows that lead to deeper engagement.
The product is in beta with acknowledged instability and active technical development (LLM spatial reasoning improvements, context engineering, agent autonomy work), but users have no visibility into progress. This creates uncertainty about whether Adam is a side project or a serious long-term platform. Users hesitate to invest time learning the tool or building workflows around it if they perceive abandonment risk.
The full product behind this analysis
Mimir doesn't just analyze — it's a complete product management workflow from feedback to shipped feature.
Themes emerge from the noise.
Ranked by severity and frequency, with the original quotes inline so you can judge for yourself.
Talk to your research.
Ask questions, get answers grounded in what your users actually said.
What's the top churn signal?
Onboarding confusion appears in 12 of 16 sources. Users describe “not knowing where to start” [Interview #3, NPS]
A prioritized backlog, not a wall of sticky notes.
Ranked by impact and effort, with the reasoning you can actually defend in a roadmap review.
PRDs, briefs, emails — on demand.
Generate documents that reference your actual research, not generic templates.
Paste, upload, or connect.
Transcripts, CSVs, PDFs, screenshots, Slack, URLs.
This analysis used public data only. Imagine what Mimir finds with your customer interviews and product analytics.
Try with your data