At Digital Sundai we are thrilled to announce the launch of our in-depth Conversational AI White Paper (63 pages!). This white paper offers a deep dive into the business strategies for deploying Conversational AI, step-by-step guidance on executing use cases, and a thorough look at the technical challenges we’ve overcome to make it all work.
No time to read 63 pages? -> We uploaded the white paper to Google NotebookLM and asked it to make a podcast. That’s all we did.
The result: an absolutely mind-blowing 15-minute podcast where ‘real people’ discuss the key insights from our white paper. Available on Spotify now.
In the rapidly evolving landscape of Artificial Intelligence, there’s significant concern about the potential risks and unintended consequences of AI technologies. From fears of job displacement and data privacy breaches to the possibility of biased decision-making and loss of human oversight, it’s evident that building trust in AI is paramount.
Trust is not just a nice-to-have; it’s a critical component for the successful adoption and long-term viability of AI solutions.
Recently, we collaborated with Mooncake AI on a project for a Municipality, focusing on responsible AI. Here are my key findings:
1. Governance & Processes
Clear ethical guidelines are only the starting point. The key is to embed these guidelines in a governance structure that seamlessly connects ethics to the different project development stages, from idea to implementation of AI solutions. This approach should be light-touch where possible (e.g., at the idea phase) and thorough where necessary (e.g., for high-risk implementations).
2. AI Capability
A minimum requirement for responsible AI is confidence in the AI capabilities of the organization, whether these capabilities are internal, with an external partner, or a mix of both. Are these experts in control of their work? Many organizations are still building this expertise. Start with simple and low-risk cases and take on more complex, higher-risk cases as your capability grows. But do grow your AI capability!
3. AI Training & Awareness
User Awareness: Educate end-users about AI systems, their capabilities, and limitations, fostering informed and responsible use of AI technologies. Build their capability to work with AI, for example, by developing prompt engineering skills.
Interdisciplinary Approach: Encourage collaboration between technologists, ethicists, and domain experts to create well-rounded AI solutions.
Continuous Education: Invest in ongoing training for AI practitioners both on the latest ethical guidelines, biases, and the social impacts of AI, as well as on their technical AI development skills.
During the project, it was encouraging to see a gradually growing trust in the organization’s capability to implement AI responsibly.
Together, let’s shape a future where AI works for everyone.
We’re thrilled to announce a groundbreaking new partnership with AI-Applied in the realm of GenAI and Text Analytics! 🤝 💥
Their profound expertise in text analytics AI, coupled with our robust Business AI consulting capabilities, positions us to explore new frontiers and drive innovation. 🌍 💡
This collaboration builds on our shared success, having previously delivered pioneering projects for industry leaders such as ING, ABN AMRO, FDMediagroep, and Bol.com. 🌟
Together, we aim to harness the full potential using AI to understand, classify, summarize and generate text & documents. Tailor-made solutions will empower our clients to digitally excel. ⚡ 🔍
We appreciate the trust our partners and clients have placed in us, and we look forward to collectively achieving new milestones and shaping the future of AI! 🚀 🙌
Mastering ChatGPT Effectively: Prompts for Business Success
By Gülbike Mirzaoğlu | Jul 10, 2023
Businesses today are constantly seeking innovative ways to improve their operations and enhance client experiences in the rapidly evolving digital landscape. ChatGPT, an advanced AI language model, offers game-changing opportunities for achieving these goals. It has the potential to revolutionize consumer interactions, streamline processes, and uncover valuable insights. With its exceptional functionality,it can enhance automation, speed up procedures, and may be a useful tool for both customers and businesses. However, ChatGPT’s shortcomings in accurately carrying out some tasks and propensity to produce erroneous or imaginary responses have drawn significant attention and spurred discussions. Therefore, it is crucial to master the art of crafting well-designed powerful prompts if one is to fully realize ChatGPT’s potential.
Getting the most out of ChatGPT
Prompting refers to the practice of formulating precise and tailored instructions or queries that guide ChatGPT’s responses. By utilizing carefully constructed prompts, users can effectively navigate the chatbot and unlock personalized and relevant responses, maximizing the benefits of ChatGPT for their specific needs.
The importance of mastering the skill of creating ideal prompts has gained increasing recognition, even leading to the emergence of a new and exciting role in the job market: AI prompt engineering. Companies are actively seeking experts who possess the ability to communicate effectively with chatbots, offering competitive salaries that can reach up to $300,000 (€275,346). This significant demand for prompt engineering professionals underscores the growing recognition of the impact that strategic prompting can have on leveraging the full potential of ChatGPT.
In this post, we will examine some of the important methods and key cues that optimize the advantages of ChatGPT for your company. By creating effective prompts, your company can improve customer interactions, streamline operations, and drive growth.
1) Be Clear and Specific: What do you want to achieve through prompting, what is your goal? When crafting prompts, it is crucial to be precise, specific, and provide explicit instructions. Specify the desired information or task, give the specifics in a detailed way. For example:
Vague Prompt:“Discuss our marketing campaign.”
Improved Prompt: “Provide an analysis of our current marketing campaign’s performance, including key metrics, ROI, and any adjustments we should consider for better results.”
2) Know your audience: It is important to understand your target audience and their preferences to tailor prompts to match the language and tone that resonates with your customers or clients. This data may include demographics, interests, behavior patterns, purchasing habits, customer segments, and pain points. Utilize surveys, interviews, customer feedback, and analytics tools to gain valuable information. For example:
Vague Prompt:“Make a social media post on our promotion.”
Improved Prompt: “Develop a social media post highlighting our latest promotion with an engaging and conversational tone. Our target audience is young professionals interested in sustainable fashion.”
3) Context-rich prompts: Context is the foundation of effective prompting. Instead of using vague or general questions, provide context-rich prompts with relevant information such as the customer’s history, preferences, or previous interactions. The better ChatGPT understands the context, the more accurate and helpful its responses will be, enhancing customer satisfaction and engagement. For example:
Vague Prompt:“Write a blog post about our company’s achievements.”
Improved Prompt: “Craft a compelling blog post highlighting our company’s recent achievements in reducing carbon emissions, achieving record sales growth, and winning industry awards. Include specific data and success stories to showcase our positive impact and position us as an industry leader.”
4) Include constraints or limitations: When crafting prompts, it’s crucial to include any specific constraints or limitations that should be taken into account. Whether it’s budgetary restrictions, regulatory requirements, or other considerations, mentioning these factors in the prompt helps guide ChatGPT to deliver relevant and feasible responses. For example:
Vague Prompt:“Develop a new product concept.”
Improved Prompt: “Develop a new product concept that aligns with our target market’s preferences and meets our budget constraints of $20,000. Consider regulatory requirements, sustainability considerations, and the need for scalability in production.”
Vague Prompt: “Design a new website layout.”
Clear Prompt: “Design a new website layout that improves user experience, adheres to our brand guidelines, and is mobile-responsive. Keep in mind the project timeline of 2 months and the need for easy navigation.”
5) Language and instruction: When engaging with ChatGPT, it’s essential to use clear instructions, strong and expressive language, and positive phrasing. Avoid corporate jargon and technical terms that might be unclear or confusing. For example:
Instead of “do not write informally”, opt for “write formally”.
Rather than “rewrite the text” use “clarify the text”.
Instead of “rewrite the paragraph,” say “revise the paragraph for clarity.”
Instead of “try not to be vague,” say “provide specific details for accurate responses.”
In interactions with an AI bot, please and thank you are unnecessary. Focus on providing precise instructions to achieve the best results.
So use clear instructions.
6) Tone and style: The tone and style of your prompts are instrumental in captivating your audience and reflecting your brand’s personality. By specifying the appropriate tone and style, you align your communication with your brand identity and cater to your audience’s preferences. Consider whether they favor formal or informal language, technical terms, or conversational phrases. To ensure clarity and consistency, it’s essential to specify the tone within your prompts. For instance, include phrases like “Tone: Informative” or “Write using a persuasive tone.” This precise instruction guides ChatGPT’s responses to match your intended style accurately. The tone options are diverse, ranging from narrative and firm to casual, informative, persuasive, gender-neutral, and more.
Another exciting approach to engaging ChatGPT is to adopt various roles and personas for your prompts. Think of your AI language model as an all-in-one team of experts! Act as a marketer to create compelling copy, a project manager to outline tasks, a best-selling author to craft captivating stories, an analyst to provide data-driven insights, or even a therapist to offer empathetic support. For example:
Vague prompt: “Respond to a customer complaint.”
Improved prompt: “Craft a polite and empathetic response to a customer complaint about a delayed delivery providing a clear explanation and offering a solution or compensation, if necessary.”
7) Define the format: Specify your preferred format or structure within the prompt to guide ChatGPT accurately. You can highlight markdowns, character limits, paragraph details, and output structure. For example:
Sales Pitch Prompt: “Craft a persuasive sales pitch with a headline, a captivating introduction, key product highlights, and a strong call-to-action.”
Markdown Formatting Prompt: “Format the response using Markdown to ensure clear headings, bullet points, and emphasized text.”
Character Limit Prompt: “Craft a concise response, limiting it to 200 characters or less.”
Tabular Output Prompt: “Conduct a comparative analysis of three competing products, listing their pros and cons in a table format with the following columns: [Column X], [Column Y], [Column Z]”
Moreover, you can create your prompts in three distinct formats to guide ChatGPT more effectively. These formats are designed to provide clear instructions and context, helping the model generate responses that align with your desired outcomes. These are:
RTF (Role, Task, Format):In this format, you start by assuming a specific role for ChatGPT. You instruct the AI about the task it should perform and specify the preferred format or style of the response. For example: “You are a marketing manager. Craft a compelling blog post promoting our latest product launch. Ensure the content is engaging and informative.”
CTF (Context, Task, Format):In this structure, you begin by providing contextual information to ChatGPT. This context helps the AI language model understand the situation better, enabling it to generate responses that are contextually appropriate. You then specify the task it should perform and highlight the preferred format, which may include specific instructions, language preferences, or word count limitations. For example: “Our company is hosting a virtual event. Create an attention-grabbing social media post to invite participants. This should be engaging, use UK English, and have a word count between 500-700 words.”
RASCEF (Role, Action, Steps, Context, Examples, Format): This format is more comprehensive and includes multiple components. You begin by assigning a role to ChatGPT, defining the specific action it needs to take, and providing step-by-step instructions to guide its response. You can also offer relevant examples and additional context, ultimately specifying the preferred format or structure of the output. For example: “You are a content strategist. Assist me in creating a comprehensive content marketing plan. Identify the target audience, select suitable social media platforms, define campaign objectives, create a content schedule, and devise a plan for result analysis. This might include email newsletters and blog posts.”
Other Tips
Reverse Prompting:
Another valuable technique in effective prompting is Reverse Prompting. With this approach, you can create a prompt based on a given text or code by extracting relevant information from the existing content. Therefore, you can create more precise and tailored prompts in line with your specific needs. Here’s a step-by-step process:
Select Text/Code: Choose the text or code from which you want to derive the prompt. Identify the key elements you wish to use for instructing ChatGPT.
Create the prompt: Use the selected text as inspiration to craft a clear and concise prompt that provides the necessary context and guidance for the AI model. For this, you can create a prompt such as: “Enable Reverse Prompt Engineering. By reverse prompt engineering I mean creating a prompt from a given text. Create a reverse prompt engineering template from this text: [text]”
Refine the Prompt: Fine-tune the prompt by adjusting the context, tone, style, and other parameters to ensure a more focused and tailored response.
Test and Iterate: Experiment with different variations of the prompt to see which one elicits the most relevant and accurate responses from ChatGPT. Continuously refine and iterate until you achieve the desired outcome.
Priming prompts:Priming prompts offer a powerful way to guide AI language models like ChatGPT and obtain different types of responses. Depending on the prompt structure, you can enable zero-shot, one-shot, or multiple-shot interactions. Let’s explore each approach in detail:
Zero-shot Priming:In this zero-shot approach, you simply provide a general instruction to the AI without specifying any additional context. The AI, equipped with its vast knowledge, will generate a social media post on the given topic without any prior information. This technique allows for creative and spontaneous responses from the AI, making it an excellent choice for open-ended explorations.
Prompt: “Write me a social media post about [topic].”
One-shot Priming: With one-shot priming, you supply a single example along with the instruction. The AI then uses this example as a reference point to create a social media post on the specified topic. This approach can yield responses that align closely with the provided example, ensuring more consistent outputs.
Prompt: “Write me a social media post about [topic]. Here is an [example].”
Multiple-shot Priming:In the multiple-shot priming method, you present several examples along with the instruction. The AI considers these examples to generate a diverse range of social media posts on the given topic. This technique encourages varied outputs, making it suitable for exploring multiple perspectives and ideas.
Prompt: “Write me a social media post about [topic]. Here are four examples: [example 1] [example 2] [example 3] [example 4]”
Temperature checks: Adjusting the temperature scale is a powerful tool to control the creativity and randomness of ChatGPT’s responses. The scale ranges from 0 to 1, with 0 representing a deterministic, focused, and conservative output, while 1 offers diverse, creative, and unexpected answers. By adjusting the temperature, you can fine-tune ChatGPT’s responses to meet your specific requirements.
Prompt: “Write a social media post about our new product launch. Temperature: [x]”
Here are the three temperature ranges and their corresponding use cases:
Lower Temperature (0.1 – 0.4): When precision is important, lower temperatures are your go-to choice. This range is useful for factual information, precise answers, or adhering to specific formats or brand guidelines.
Moderate Temperature (0.5 – 0.7): This range offers achieving a balance between creativity and consistency. Balances creativity and consistency. This setting is especially suitable for general content generation, combining accuracy and inventiveness, for example in crafting blog posts, email newsletters, or product descriptions.
Higher Temperature (0.8 – 1.0): If you are seeking to break the boundaries of conventional thinking, this higher temperature range is your gateway to limitless creativity. When brainstorming marketing campaigns, crafting engaging social media content, or exploring fresh perspectives, this range encourages ChatGPT to produce diverse and imaginative responses, unveiling novel ideas and unique perspectives.
A Work In Progress: Understanding Limitations
ChatGPT’s capacity to revolutionize business operations is undoubtedly exciting. However, it is crucial to understand its limitations to ensure ethical and secure application.
ChatGPT’s proficiency heavily relies on context, which can sometimes lead to limitations in its responses. OpenAI acknowledges that the technology lacks knowledge of events beyond its data set’s cutoff date (September 2021) and does not learn from its prior experiences. This may result in what is known as “hallucination,” where the AI generates content that may sound plausible but lacks grounding in reality. This can be concerning as it may produce harmful advice, buggy code, or inaccurate information.
Additionally, ChatGPT might occasionally make “simple reasoning errors” or accept obvious false statements without cross-checking, making it important to be mindful of such occurrences. While it excels at specific tasks, comprehending real-world nuances can be challenging. Being mindful of these aspects enables you to interpret ChatGPT’s responses thoughtfully and make well-informed decisions.
Prioritizing Data Privacy
Data privacy is of utmost importance when using AI technologies like OpenAI. To maximize the benefits of ChatGPT for your business, it’s crucial to be cautious and remain well-informed about the privacy policies in place.
However, it’s essential to note that using ChatGPT without the API entails differences in security practices. Whether you opt for the free GPT-3.5 version or the paid ChatGPT Plus with access to GPT-4, ChatGPT automatically collects certain personal information, such as log data, usage data, device information, and cookies. Additionally, it may gather personal information that users provide, such as account details, user content, communication information, and social media data.
Given these data collection practices, if you do not integrate ChatGPT through the API, safeguarding your business and customer information should be a top priority. In this case, it’s crucial to take necessary precautions to protect sensitive data and ensure compliance with privacy regulations. By being mindful of the data shared with AI technologies like ChatGPT, businesses can foster trust with their customers and maintain a strong commitment to data privacy and security. Here are some recommended precautions to safeguard your privacy:
Avoid Sharing Sensitive Information: When interacting with ChatGPT, avoid sharing sensitive business or personal details to ensure data confidentiality.
Anonymize Data: Whenever possible, anonymize or aggregate data before using it with ChatGPT to reduce the risk of identifying individuals.
Use a VPN: Implementing a Virtual Private Network (VPN) can add an extra layer of protection, which can secure your online identity and data from potential vulnerabilities.
Opt-out of Personal Data Processing: It is important to check whether ChatGPT allows users to opt-out of personal data processing. This can provide more control over the information shared with the AI.
Caution with Third-Party Apps and Plug-ins: While integrating third party apps and plugins with ChatGPT can be useful for your business, it is important to be mindful that they may access your data without your knowledge. It is important to choose trusted sources to maintain data integrity.
Review Data Policies and Stay Updated: Familiarize yourself with OpenAI’s data policies, terms of service, and privacy agreements to ensure compliance. Stay informed about any changes to OpenAI’s privacy policies or security measures and adapt accordingly.
Conclusion
ChatGPT presents game-changing opportunities for businesses to revolutionize interactions, streamline processes, and uncover valuable insights. However, acknowledging its limitations and ensuring ethical usage is crucial. To fully realize ChatGPT’s potential, crafting powerful prompts is essential.
This post explored key methods for optimizing ChatGPT’s advantages, including clear and specific prompts, context-rich instructions, and defining response formats. Additionally, it addressed data privacy concerns and highlighted measures to safeguard sensitive information. By embracing ChatGPT responsibly and continuously adapting prompts, businesses can unleash its transformative capabilities for sustained success in the digital age.
Artificial intelligence (AI) is becoming increasingly important in our lives and has undergone some remarkable advancements. One of the earliest forms of AI is traditional AI, also known as conventional AI. It is capable of describing, predicting, and prescribing based on existing data. This has proven to be quite useful in a variety of areas, such as recognizing patterns, predicting client churn, forecasting product demand, providing next-best product recommendations, speech and image recognition, and natural language processing. In other words, traditional AI can analyze and understand patterns and data, and provide valuable insights and recommendations that can help businesses make better decisions.
What Are The Creative Possibilities of Generative AI?
Generative AI, on the other hand, is a type of AI that has become a game changerin its ability to generate new content. It is based on deep learning algorithms and uses neural networks to learn from large datasets, taking the creativity to the next level. Generative AI can create new images, audio, video, coding, 3D object generations, and even text that is indistinguishable from human-generated content. It is capable of generating numerous contents including but not limited to new music tracks, sound effects, code/dataset generations, content writings, solving exam questions, chatbots. Despite being in the early stages of development, the outputs generated by generative AI have already been incredibly impressive. For example, in 2018, a piece of artwork called “Portrait of Edmond de Belamy” was generated using an AI algorithm (Figure 1) andsold for $432,500. Similarly, in the same year, fashion brand Carlings introduced a “digital collection,”which consisted of clothing designs that were completely created using generative AI.
Figure 1: Portrait of Edmond de Belamy
Numerous companies and organizations offer generative AI solutions and services. Among the most renowned are OpenAI, which has created language models such as the GPT series, DALL-E 2, JukeBox, and Point-E. Microsoft provides VALL-E, RODIN Diffuson, GODIVA, and MoLer. IBM has the Watson AI platform, Google offers DeepDream and Magneta, while Amazon provides Lex and DeepComposer, and NVIDIA has MT-NLG, Edify, and MegaMolBART.
Meet ChatGPT: The Best AI Chatbot of 2022
Among many other models, ChatGPT has become the most popular and even considered the best AI chatbot ever since it became available to the public in November 2022. Within just in a week, it had attracted over a million users. To give you an idea of how impressive this is, reaching one million users took 3.5 years for Netflix, two years for Twitter, ten months for Facebook, and 2.5 months for Instagram (Figure 2).
ChatGPT is a large language model (LLM), which is a type of generative artificial intelligence model that is trained to understand and generate natural-sounding human language. It is based on the GPT (Generative Pre-trained Transformer), which is one of the most advanced deep learning architectures for natural language processing (NLP). Its highly advanced and sophisticated language model is capable of generating human-like responses to a wide range of questions and queries.
ChatGPT is a powerful tool that can understand questions and provide relevant and informative responses. It has been trained on a vast dataset of text, including books, articles, and websites, which gives it a broad range of knowledge and information to draw from. ChatGPT is versatile and can be used for various applications like virtual assistants, customer service chatbots, and educational tools. It can handle multiple languages and can be customized to specific domains and topics. Additionally, it is easily accessible as an API (Application Programming Interface), allowing developers and organizations to integrate it into their own applications and services. This has made it simpler for businesses and individuals to use ChatGPT to enhance their customer service, automate repetitive tasks, and provide personalized recommendations to users.
Figure 2: This chart shows the time it took for selected online services to reach one million users.
What Are The Risks and Challenges of Generative AI?
The use of Generative AIs has transformed the way we interact with and perceive AI. However, there are also significant risks associated with it that require continuous adaptation and change from AI developers, business users, investors, policymakers, and citizens. These risks include trust, security, and governance concerns such as data breaches, adversarial attacks, ethical considerations, deepfakes, data privacy, legal compliance, copyright issues, and cybersecurity problems. A comprehensive approach is necessary to address these risks, including a clearly defined strategy, good governance, and a commitment to responsible AI. Any organization that embraces generative AI must consider these issues with ethics in mind.
Impact of Generative AI on Business
It has become evident that the implementation of generative AI will have a significant impact on various business functions in the long run. However, in the initial stages, the greatest impact will be observed in areas such as information technology, marketing and sales, customer service, and product development. We also expect a significant impact on legal and risk. These functions will witness a remarkable transformation in terms of efficiency, accuracy, and productivity, expected to lead to a net positive impact on the overall business performance.
Generative AI is set to revolutionize various business functions in the foreseeable future. Initially, its greatest impact will be felt in key areas such as information technology, marketing and sales, customer service, and product development. As technology advances, we can expect to see it used more widely across various industries. These developments are expected to have a positive impact on overall business performance, ultimately driving growth, efficiency, accuracy, and success.
Customer Service: Natural-sounding, automated, personalized chatbots, virtual assistants. With the help of AI, customer interactions can be more consistent and aligned with the desired tone and messaging. AI systems can provide knowledge bases and recommendations to human agents, enabling them to resolve customer issues faster and more effectively. This helps to enhance human agents’ capabilities rather than replacing them entirely.
Marketing & Sales: Generative AI applications can be used for various tasks such as SEO optimization, creating ads, personalized messaging, detecting fraud, recommending new products, services, or offers. These applications can also generate content for customer outreach, including blog posts, social media posts, product descriptions, website copy, and more. It is anticipated that within two years, generative AI systems will assist in developing 30% of all outbound marketing messages.
Information Technology: Artificial intelligence models that are designed to generate output have a vast array of uses within the realm of information technology. They are capable of creating codes, detecting and resolving bugs within existing codebases, addressing common IT problems that arise for both employees and end users, optimizing network performance, automating the setup and management of infrastructure, assisting with cybersecurity efforts, automating tasks that are repetitive in nature, and analyzing vast quantities of data to produce valuable insights and recommendations for IT systems. Automated coders currently available on the market have increased developer productivity by over 50%, which has contributed to the speedy progress of software development.
Product Development: Generative AI is a useful tool for companies to come up with innovative product and feature ideas that may not have been considered by humans. By analyzing a vast collection of patents, generative AI can identify opportunities and risks related to intellectual property and new product features. It can also aid designers by producing design concepts, 3D models, color schemes, and materials that meet specific criteria, which accelerates the design process. Additionally, generative AI can facilitate faster prototype iteration for product designs.
What’s Next? Implications and Possibilities of the Limitless Potential of Generative AI
The potential of generative AI seems nearly limitless. Generative AI is a rapidly evolving technology that enables machines to produce creative works such as text, images, and audio that are almost indistinguishable from those created by humans. Prominent examples of this technology include DALL·E, GPT-3, and Stable Diffusion, which have demonstrated remarkable capacities for generating content on a massive scale. Given these developments, it is reasonable to wonder about the potential implications and possibilities for this transformative technology in the future.
As the technology behind Generative AI models continues to advance, their potential applications are becoming increasingly extensive and their outputs more human-like. These models have the potential to offer greater levels of personalization, collaboration, conversation, and interactivity. With the continued improvements in machine learning, goal-directed learning, perception, and planning, it is expected that AI will evolve from a generic tool to a truly autonomous intelligent agent. These agents will be able to perceive, comprehend, set objectives, and act independently, ushering in a new era of AI capability and functionality.
Proudly announcing the Whitepaper on the Use of AI. A research by Erasmus Centre for Data Analytics, Kenniscentrum Business Innovation and Digital Sundai
Download now and you will get insight in:
How organizations are using AI today, and the key trends.
What the business impact is of AI. Today and in the near future.
AI/ML enables the creation of superior performing organizations. What these digital and AI/ML enabled organizations exactly look like is unknown. Innovation, creativity, perseverance and digital business & AI/ML technology craftsmanship will create them. We are looking for a senior big data / ML engineer that will push the boundaries for our customers.
Your role
Your role as Senior Big Data/ML Engineer is to demonstrate clients the art-of-the-possible and apply AI/ML technologies to create business benefits. You will be working with clients to define their AI and Data roadmaps, improve their capability, but foremost you will be realizing and scaling their state-of-the-art cases. Use cases could be for example the introduction of voice enabled chatbots, text analytics to classify emails or documents, or a digital twin to reduce waste of a production process.
Typically you will apply data collection, scaling AI/ML models and building data pipelines while working in multidisciplinary teams. The Google Cloud Platform provides our preferred technology stack. Essential for success will be your seamless cooperation with the digital business consultants, data scientists and software engineers from Digital Sundai, our network partners and clients. Next to being a great engineer we expect you to be a strategic innovator of our services.
Where
At Digital Sundai based out of Amsterdam. A young and ambitious start-up.
Digital Sundai strives to create superior organizations through Digital & AI. We believe Digital & AI projects only succeed when technology & business are both done right. We bring experienced digital business competence and top AI & Analytics expertise. Executed through our agile digital culture and methodology. Digital Sundai is a networked enterprise which only works with top digital talent & top digital partners. We are an Open company and an integral part of the Digital community with relations and access to the latest Business & Tech start-ups, scale-ups, academia, and established companies. Google Cloud is our preferred Technology ecosystem
Working in creative, agile teams, you’ll help clients discover, design and unlock the digital opportunity for their organizations. You’ll help them modernize their data and analytics environments, and apply next generation AI/ML technologies. These are technologies driven by the transformational impact of AI-enabled, automated processes and optimized through human-centered design.
What we offer
a fixed salary.
employee ownership of the company (shares) as we believe in entrepreneurship.
flexible working hours and the opportunity to work from home.
a good pension scheme.
25 days of paid holiday annually, and the opportunity to purchase additional holiday days annually.
a 32- 40 hour working week.
the opportunity to take unpaid leave.
a good mobility policy allowing you to lease a car, use public transport or get reimbursed for travel costs.
a laptop and a phone, which are also for personal use.
great growth opportunities. Depending on your ambitions and performance, you can grow your impact very fast, and reap the benefits as co-owner of the company.
What you offer
Academic Master Degree, preferably in Information Sciences, Data Science or Analytics. Additional experience at digital service providers is making you an even better fit.
>4-6 years of relevant experience as a Big Data / ML engineer with a track record of many successful AI/ML & Analytics implementations.
Digital technology savvy and with a passion for AI/ML & engineering. And you believe as us that this is the ultimate professional playground for years to come.
Ambition to gain Expert knowledge in leading AI & analytics tools such as Tensorflow, Hadoop, NoSQL, Kubeflow, and modern SW technology like docker and CI/CD tools. Good working knowledge of Python and R.
GCP certification like Professional Data Engineer.
Experience in supervised, unsupervised and reinforcement learning approaches, designing and building data processing systems, designing, building and operating machine learning models and ensuring solution quality.
A digital business mindset
You are ambitious, curious and entrepreneurial
You are a team player and you have a proven record functioning in multidisciplinary teams
