# Dashbouquet > Dashbouquet is a reviewed AI agency in Tallinn, Estonia, listed in the AIAgencies.eu register and last assessed 2026-09-03. Canonical page: https://www.aiagencies.eu/agency/dashbouquet Generated: 2026-09-28 · Curated by Gabor Kiss · No paid placements — inclusion, score, and rank are never sold. ## Facts - Location: Tallinn, Estonia - Website: https://dashbouquet.com - Hourly rate: €45-85/hr (directory-listed) - Team size: 50-249 - Founded: 2014 - Services: AI Development, Generative AI, AI Consulting - Industries: Healthcare, SaaS & B2B, E-commerce - LinkedIn: https://www.linkedin.com/company/dash-bouquet ## About Product engineering firm building LLM-backed applications — conversational assistants, multimodal analysis and AI-driven data platforms — alongside web and mobile development. Registered in Tallinn with a Warsaw office. ## Documented clients Documented engagements: Dashmote, BetterYou, Pyroscope, utheory, Nubabi, VMware, and Autodesk. 5 of 7 carry a source URL on file. ## Curator assessment Reviewed 2026-09-03 by Gabor Kiss against published evidence. The AI evidence splits into two kinds: production-scale data extraction—Dashmote's beverage-menu platform pulled menus from 10,000-plus venues across 45-plus countries in 12 languages, processing 100 GB a day on AWS, Kubernetes, Argo Workflows, SpaCy and PyTorch over 16 months—and early LLM product work, where BetterYou's therapist dashboard and a dating-assistant app document model bake-offs (GPT-4, Claude 3, Llama 3, Mixtral) but publish no usage or accuracy figures. That split is a buyer consequence: the extraction and BI systems (a second one tracked 21,000 SKUs across 9 marketplaces for 18 months) show a team that keeps scrapers and parsers alive against layout changes, while the LLM studies read as pre-product builds, so if your brief is a generative feature in production, ask what happened after launch. The strongest outcome on the site is not an AI one—Pyroscope credits the firm with a 2x revenue lift and 30% more daily active users before its acquisition by Grafana Labs—which counts as a delivery signal and not as an AI signal. The weaker fit is buyers who need a model team rather than a product team—the published work puts LLMs and OCR inside apps the firm also builds, with no study where the model itself was the deliverable. ### Key strengths - Team composition and duration are published per data project (2 engineers, 2 analysts, 1 PM over 16 and 18 months), so you can size a comparable engagement - Sprint rhythm is stated and consistent across the site: features every 1–3 weeks, two-week demos, and a Clutch 1000 listing for 2025 - Pyroscope's CEO reviews the firm by name on Clutch, and the collaboration is described down to the Node.js and Ruby profiling agents delivered ### Good to know - Ask what became of the BetterYou and dating-assistant builds after handover—both studies stop at model selection with no adoption or accuracy numbers - Confirm who leads the ML work: the about page names no engineers, and the AI services page lists Caffe, Theano and CNTK among its platforms, which are years out of maintenance - The AI services page positions the firm around OpenAI-style APIs—if you need a fine-tuned or self-hosted model, ask for a reference beyond the SpaCy and OCR extraction work ## Compared head-to-head - Dashbouquet vs DATAFOREST: https://www.aiagencies.eu/compare/dashbouquet-vs-dataforest - Dashbouquet vs MindTitan: https://www.aiagencies.eu/compare/dashbouquet-vs-mindtitan --- 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