# craftworks vs enliteAI

> Two top-rated AI Development agencies in Vienna. Side by side: published rates, team size, service and industry coverage, and the curator's assessment of both — pairings are never sold.

Canonical page: https://www.aiagencies.eu/compare/craftworks-vs-enliteai
Generated: 2026-09-17 · Curated by Gabor Kiss · No paid placements — inclusion, score, and rank are never sold.

## Facts
| | craftworks | enliteAI |
|---|---|---|
| Editorial score | 61.8/100 | 52.8/100 |
| Hourly rate | not published | not published |
| Team size | 10-49 | 10-49 |
| Location | Vienna, Austria | Vienna, Austria |
| Founded | 2014 | 2017 |
| Last reviewed | 2026-09-07 | 2026-09-07 |
| Services | AI Development, Generative AI | AI Development, AI Agents, AI Consulting |
| Industries | Manufacturing, Healthcare | Public Sector, Manufacturing |

## Our take on craftworks
Reviewed 2026-09-07 by Gabor Kiss against published evidence.

Delivery evidence sits in the navio product line—published success stories describe visual-inspection systems running on production lines with measured outcomes, and navio itself is an MLOps platform with model retraining, blue/green deployment, out-of-distribution detection and edge rollout to Azure IoT and TTTech Nerve.

Names and numbers sit on different pages: VERBUND, Wien Energie, voestalpine and ZKW Group are quoted by name and role in testimonials, while the recent success stories describe "a leading industrial manufacturer" or "a large energy plant operator".

The team page lists 38 people including named engineers, and the AI-sovereignty page addresses on-premise deployment, the EU AI Act and NIS2 directly rather than as a footnote.

The weaker fit is buyers outside industrial operations, or anyone who needs the model architecture named before signing—published detail on the vision work stops at "AI-powered models".

### Key strengths
- Published outcomes carry numbers: 94% accuracy classifying bearings with 82% precision on defect type, 72% faster package inspection, 60% less time on manual toolkit checks
- navio covers the operational side most AI suppliers leave out—retraining, shadow deployments, data-drift and out-of-distribution monitoring, with deployment to edge hardware
- The VERBUND engagement is documented as a multi-year rollout, extended to the Drau and Malta power plant groups, linking their AKS measurement database to their GAMS optimization engine

### Good to know
- The success stories with the strongest numbers are anonymized—ask which named client each deployment belongs to and for a reference you can speak to
- No rates published; anchor against the European range on our rate index and clarify whether navio is licensed separately from the delivery work
- Generative AI is listed as a service but no success story documents a generative deployment—ask for one with an outcome if that is your project

## Our take on enliteAI
Reviewed 2026-09-07 by Gabor Kiss against published evidence.

The published evidence is research and product, not client delivery: enliteAI maintains Maze, an open-source applied reinforcement-learning framework, co-authored a 2025 paper with grid operator TenneT on power-grid topology control, and spun out Detekt, a GeoAI road-asset platform acquired by Cyclomedia in August 2025.

Fifteen client logos are published—A1, Andritz, Audi, Boehringer Ingelheim, EY, Erste Group, voestalpine and the City of Vienna among them—but no engagement is described anywhere on the site, so what was built for whom stays unstated apart from the 2019 municipal AI strategy for Vienna.

The energy work is presented as a methodology and a maturity model—a seven-step path toward autonomous digital twins for distribution grids—with integration claimed for Schneider Electric, Hitachi, Siemens and GE Vernova systems.

Best for grid operators and optimization problems where reinforcement learning is genuinely the method; a buyer who needs to read what a comparable engagement delivered will have to ask for it directly.

### Key strengths
- The method is checkable in public: a 2025 paper co-authored with grid operator TenneT on multi-objective grid topology control, plus first place in the 2022 L2RPN power-grid competition
- Maze, their applied reinforcement-learning framework, is on GitHub—you can read how they build agents before hiring them
- Detekt, the GeoAI platform they built and spun out, was acquired by Cyclomedia in 2025, which is shipped-product evidence even though it was their own venture

### Good to know
- No client case study is published anywhere on the site—ask for a written engagement summary with outcomes before shortlisting
- The energy work reads as a methodology and pilot framework rather than a live deployment—ask which grid operator is running it and at which maturity step
- Computer vision is still listed as a capability, but Detekt, the team behind it, was sold in August 2025—confirm what vision capacity remains in-house

## Which to choose
Choose craftworks for:
- disclosed specialization in Healthcare
- broader service offering — also covers Generative AI
- a higher overall editorial score

Choose enliteAI for:
- disclosed specialization in Public Sector
- broader service offering — also covers AI Agents and AI Consulting

## Questions buyers ask

### Which has more Manufacturing experience, craftworks or enliteAI?
Both agencies show documented Manufacturing work and have similar industry breadth. Compare directly on the agency profiles: craftworks and enliteAI.

### Which scores higher overall, craftworks or enliteAI?
craftworks scores 61.8/100, 9 points higher than enliteAI at 52.8/100. Both portfolio quality and business credibility feed into the score — see our methodology for the full breakdown.

### When should I consider both craftworks and enliteAI?
Consider running parallel discovery briefs with craftworks and enliteAI if your project spans multiple workstreams, you want competitive proposals to compare scope and approach, or you're undecided between the specialization angles each brings (see the Best for cards above). Most engagements ultimately go with one — but the parallel-brief phase is a low-cost way to validate fit.

---
Source: AIAgencies.eu — the curated register of European AI agencies. Rate provenance: "agency-confirmed" = disclosed via submission or claimed profile; "directory-listed" = Clutch or two agreeing B2B directories (lone unverified sources are never written). Full method: https://www.aiagencies.eu/methodology · Rate index: https://www.aiagencies.eu/rates