SpinningBytes
SpinningBytes is a Swiss NLP and speech-processing company, founded 2015 as an ETH/ZHAW spin-off, that builds custom AI solutions and the Interscriber transcription product for European enterprise and public-sector customers working with German and Swiss German data.
- Company typePrivate
- Founded2015
- HeadquartersWinterthur, Switzerland
- Headcount11–50
- GTM typeB2B
- OfferingSoftware
What SpinningBytes does
SpinningBytes AG is a privately held Swiss company, founded in February 2015 as a joint spin-off of ETH Zurich and the Zurich University of Applied Sciences (ZHAW), headquartered in Winterthur. The firm develops natural language processing (NLP) and speech processing solutions based on machine learning and deep learning, including the use of large language models such as GPT, Falcon, LLaMA, and Vicuna. Its core technical assets include proprietary corpora for German and Swiss German (SB-CH, SB-10k, Swiss Twitter Corpus, CEASR), an ensemble speech-to-text product (Interscriber) that combines five processing engines and supports on-premise Docker deployment, and a multi-language sentiment library. The company has won multiple international NLP competitions (SemEval 2016, Evalita 2016, GermEval 2020) and is the main sponsor of the SwissText conference.
SpinningBytes serves enterprise and public-sector customers who need to extract structured information from text and audio in regulated or low-resource language contexts — notably Swiss German dialects. Named clients include SwissRe, Beekeeper, Supertext, University Hospital Zurich, Tamedia, Job Cloud, World Vision Switzerland, MP Technology, and ZHAW. The company operates primarily in Switzerland with a stated European scope.
The business model is sales-led and combines two revenue streams. The first is bespoke professional services: brainstorming workshops (2–4 hours), rapid prototypes (5–15 days), production-ready solutions with REST/Docker delivery, data labeling, and NLP training. The second is software: one-time licenses for libraries such as the Sentiment Library (English, German, Spanish, Italian, Dutch) and recurring Interscriber subscriptions. Pricing is quote-based and undisclosed publicly; the company reports no external funding rounds and remains founder-owned under CEO Mark Cieliebak.
SpinningBytes firmographics
Firmographics- Name
- SpinningBytes
- Legal name
- SpinningBytes AG
- Website
- https://spinningbytes.com
- Company type
- Private
- Founded year
- 2015
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- SpinningBytes is a Swiss NLP and speech-processing company, founded 2015 as an ETH/ZHAW spin-off, that builds custom AI solutions and the Interscriber transcription product for European enterprise and public-sector customers working with German and Swiss German data.
- Ownership category
- akta.pro rank
SpinningBytes industry classification
Industry- Product category
- Natural Language Processing Software
- NAICS
- Custom Computer Programming Services (541511), Software Publishers (5132)
- SIC
- Services-Computer Programming Services (7371), Services-Prepackaged Software (7372)
- akta.pro primary industry
- Conversational NLP & Dialogue Management (chatbots, agents) (HDAAADAF)
- akta.pro secondary industries
- Audio & Speech Analytics (call analytics, QA, insights) (HDAAAFAH), Content Generation & Writing Assistance (copy, emails, reports) (HDAAADAG), NLP Developer Tools & Evaluation (prompting, testing, benchmarks) (HDAAADAK), Summarization & Document Understanding (HDAAADAE), Machine Translation & Multilingual NLP (HDAAADAD)
Keywords
Where SpinningBytes is headquartered
LocationHeadquarters
- HQ city
- Winterthur
- HQ country
- Switzerland
- HQ region
- Europe
Offices1 record
Markets served
SpinningBytes business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Personnel, Technology or R&D, Marketing or Sales, Operations, Infrastructure
Revenue model
- Custom NLP Projects: Bespoke consulting and development projects including brainstorming workshops, rapid prototypes, production-ready solutions, data labeling, and training services. Time and cost depend on project size.
- Software Licenses: Ready-to-use software libraries and solutions such as Sentiment Library for multiple languages (English, German, Spanish, Italian, Dutch) and Interscriber for transcription.
- Interscriber Product: Automatic transcription product using combination of 5 speech processing engines with optional local server deployment.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Other | Multi-year contract | Custom project-based pricing |
Go-to-market motion1 record
Distribution channels1 record
Marketing channels4 records
SpinningBytes product offering
Product offeringCore offering
SpinningBytes develops and delivers natural language processing (NLP) and speech processing software solutions for enterprise customers, anchored by its flagship product Interscriber for automatic audio transcription and a multilingual Sentiment Library. Beyond packaged software, the company provides custom NLP development projects covering chatbots, dialogue systems, sentiment analysis, topic categorization, machine translation, and speaker recognition, typically delivered as Docker-packaged software components with REST interfaces and source code handover.
Product overview
SpinningBytes is a Swiss NLP company offering a portfolio of products and services centered on text analytics and speech processing. The core product is Interscriber, an automatic transcription tool for interviews using multiple speech processing engines. Additional products include a Sentiment Library supporting multiple languages, and custom NLP solutions covering chatbots, dialogue systems, sentiment analysis, machine translation, and more. The company also provides data resources (SB-CH, SB-10k, Swiss Twitter Corpus) and professional services including workshops, prototyping, and deployment. All solutions can be delivered as Docker-wrapped software components with REST interfaces.
Differentiator
Problem solved
Functional benefit
Brands
- Interscriber: Software for automatic transcription of interviews, combining multiple speech processing engines for improved quality. Can run on local servers for data privacy.
Products and services
- Interscriber Automatic transcription software for interviews and meetings that combines five speech processing engines to consolidate transcription results for improved quality. Supports multiple languages and offers optional local server deployment for data privacy. Targeted at enterprise customers needing reliable speech-to-text, including Swiss German.
- Sentiment Library Ready-to-use software library for sentiment analysis supporting English, German, Spanish, Italian, and Dutch. The library classifies text as positive, negative, or neutral using machine learning. Targeted at enterprise customers needing multilingual sentiment classification out of the box.
- Custom NLP Solutions Tailored NLP software solutions for enterprise customers covering chatbots, dialogue systems, sentiment analysis, topic categorization, automated customer support, author profiling, machine translation, headline generation, speaker recognition, summarization, and text-to-speech. Delivered as custom projects or production-ready software components with architecture definition and deployment in customer infrastructure.
- Professional Services Service engagement offerings including brainstorming workshops (2-4 hours), rapid prototypes (5-15 days), production-ready solution delivery, data labeling services, data analysis and reports, and training and education on NLP fundamentals and applications. Audience ranges from novices to tech experts.
- SB-CH Swiss German Sentiment Corpus Publicly available Swiss German sentiment corpus containing 165,916 Swiss German sentences, with 2,799 labeled by 5 annotators across sentiment categories. Released under Creative Commons CC BY 4.0 and created in collaboration with ZHAW for use in Swiss German NLP and sentiment analysis research and applications.
- SB-10k German Sentiment Corpus Publicly available German sentiment corpus containing 9,738 German tweets, each labeled by 3 annotators across sentiment categories. Released under Creative Commons CC BY 4.0 and created in collaboration with ZHAW for use in German-language sentiment analysis.
- Swiss Twitter Corpus Continuously collected compilation of tweets related to Switzerland containing approximately 3.6 million tweets with metadata, used for Swiss sentiment analysis and social media monitoring. Available for academic and commercial use.
Quantifiable outcome
- Won SemEval 2016 - outperformed all teams on sentiment analysis task
- +3 more outcomes
Companies that use SpinningBytes
Customer profileNamed customers11 records
Segments1 record
Ideal customer profiles1 record
SpinningBytes technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- No
- API docs
- API detail
Core technology
AI maturity
App detail
AI capability5 records
Feature8 records
SpinningBytes partnerships and signals
Strategic signalPartnerships
Six partnerships are on record, tiered core and minor.
- Swiss Federal Institute of Technology Zurich (ETH)coreAccredited spin-off of ETH Zurich. Technology developed through three research projects funded by Innosuisse/CTI. Collaboration on deep learning research including DeepText project.
- Zurich University of Applied Sciences (ZHAW)coreClose collaboration with ZHAW research team led by Prof. Mark Cieliebak. Joint spin-off of ETH and ZHAW. Active research partnership including Innosuisse projects on dialogue transcription and NLP. Shared resources including SB-CH and SB-10k sentiment corpora.
- Centre for AI at Zurich University of Applied SciencescoreCollaboration on evaluating and implementing Large Language Models (LLMs) including ChatGPT, Falcon, LLaMA, and Vicuna since 2019.
- Data Innovation AllianceminorProud member of Data Innovation Alliance. Won Startup Grant 2019 for membership benefits.
- Swiss Association for Natural Language Processing (SwissNLP)minorMember of SwissNLP which promotes NLP, computational linguistics and text analytics in Switzerland through events and projects.
- Swiss Made SoftwareminorMember of Swiss Made Software label representing Swiss values in software development and commitment to Switzerland as software development location.
Scale indicators7 records
Recent moves8 records
Expansion highlights5 records
SpinningBytes competitors and assessment
Company assessmentDirect peers
- Deepgram: Speech-to-text platform combining proprietary and foundation models, selling to enterprise call-center and media customers. Comparable as a speech-recognition specialist targeting the same transcription workloads as Interscriber.
- MeaningCloud: Text analytics platform offering sentiment, classification, and topic extraction APIs with multilingual support. Similar mid-sized specialist competing for enterprise document-understanding workloads.
- Speechmatics: UK-based automatic speech recognition company offering multilingual transcription APIs, including accent and dialect coverage. Direct comparable to SpinningBytes' Interscriber product and ASR roadmap.
- Expert.ai: Italy-based NLP/text analytics company offering sentiment analysis, categorization, and custom enterprise NLP solutions. Closely comparable in size, European base, multilingual focus (including Italian), and hybrid product-plus-services go-to-market model.
- AYLIEN: Provides text analytics APIs (sentiment, entity extraction, classification) with strong multilingual coverage. Comparable as a focused NLP/text-analytics player competing for the same enterprise use cases SpinningBytes addresses.
Emerging players
- Rasa: Open-source conversational AI / dialogue systems framework widely used by enterprises to build chatbots. Comparable on the dialogue-systems/chatbot axis of SpinningBytes' offering.
- Primer.ai: Enterprise NLP platform focused on multi-document summarization, entity extraction, and intelligence workflows. Comparable as an NLP-first vendor serving defense, financial, and corporate analysts.
- Hugging Face: Open-source NLP/AI platform hosting thousands of models including multilingual and Swiss-adjacent work. Comparable as an ecosystem player whose open models both compete with and feed SpinningBytes' development.
Broad incumbents
- IBM watsonx: IBM's enterprise AI suite covering NLP, speech-to-text, and conversational AI. Comparable as a broad incumbent offering overlapping capabilities that enterprises may select over a specialist like SpinningBytes.
- Google Cloud Natural Language / Speech-to-Text: Hyperscaler NLP and speech APIs with strong multilingual coverage and tight integration with enterprise data infrastructure. Comparable as a large incumbent whose foundation-model improvements directly pressure SpinningBytes' differentiation.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat5 records
Key risks7 records
Key highlights7 records
Customer concentration
SpinningBytes social profiles
Digital presenceSpinningBytes compliance and trust
Trust signalCompliance1 record
SpinningBytes financial estimates
Financial estimateRevenue estimate
Valuation estimate
SpinningBytes leadership team
Management profileNumber of profiles
Profiles2 records
SpinningBytes funding detail
Funding detailFunding overview
Funding rounds
Investors
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
SpinningBytes M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about SpinningBytes
What does SpinningBytes do?
SpinningBytes develops and delivers natural language processing (NLP) and speech processing software solutions for enterprise customers, anchored by its flagship product Interscriber for automatic audio transcription and a multilingual Sentiment Library. Beyond packaged software, the company provides custom NLP development projects covering chatbots, dialogue systems, sentiment analysis, topic categorization, machine translation, and speaker recognition, typically delivered as Docker-packaged software components with REST interfaces and source code handover.
Is SpinningBytes a public or private company?
SpinningBytes is a private company. It is classified as founder individual operated bootstrapped and is currently operating.
When was SpinningBytes founded?
SpinningBytes was founded in 2015. It employs 11 to 50 people.
Where is SpinningBytes based?
SpinningBytes is headquartered in Winterthur, Switzerland, in the Europe region.
How does SpinningBytes make money?
Three revenue lines are on record. Custom NLP Projects are the primary driver. The others are software Licenses and interscriber Product.
Who are SpinningBytes's main competitors?
Direct peers on record are Deepgram, MeaningCloud, Speechmatics, Expert.ai and AYLIEN. Emerging players are Rasa, Primer.ai and Hugging Face. Broad incumbents are IBM watsonx and Google Cloud Natural Language / Speech-to-Text.
Does SpinningBytes have an API?
No public API is recorded for SpinningBytes.
What industry is SpinningBytes in?
SpinningBytes's product category is Natural Language Processing Software. Its primary akta.pro industry code is HDAAADAF, Conversational NLP & Dialogue Management (chatbots, agents), with a secondary code of HDAAAFAH, Audio & Speech Analytics (call analytics, QA, insights). Its NAICS code is 541511 and its SIC code is 7371.