Business growth today depends not just on adopting new software tools, but on integrating them with data and AI technologies -mostly Open Source Software. By embedding your organization's expertise in AI models, you can not only persist and scale knowledge to maintain a competitive advantage but also free creative resources for new business models. Instead of being swayed by the noise on eg. LinkedIn and other people's ideas and successes, focus on building your own data & AI solutions.
1. AI Implementation: Moving beyond a high-level understanding of AI, we'll explore guidelines and best practices for identifying the right starting point for AI adoption in your business.
2. From Prototypes to Production: Although open source and boilerplate code make prototyping easier, we'll examine various use cases to understand how to transition from prototypes to production, and other options to consider.
3. Navigating the Journey: Learn best practices for dealing with your business's legacy systems and determine when migrating to Open Source is a viable plan.
4. Driving Impact at Scale: Innovation encompasses various elements, including cloud, open source software, software engineering, and corporate culture. We'll discuss architectural best practices for managing multiple components simultaneously.
5. Overcoming Inexperience: If you have the necessary resources but lack experience, discover how to identify missing elements and accelerate progress in your digital transformation journey.
Each of these insights will be supported by real-life use cases, demonstrating the challenges and solutions involved in maximizing AI's value for your business and maintaining a competitive edge. Furthermore, integrating AI effectively can also enhance company culture and contribute to happier, more engaged employees, creating additional value for your organization.
2. ALEXANDER CS HENDORF
ASKING AT PYDATA BELGRAD
Business growth today depends not just on adopting new
software tools, but on integrating them with data and AI
technologies -mostly Open Source Software.
3. ALEXANDER CS HENDORF
ASKING AT PYDATA BELGRAD
Business growth today depends not just on ADOPTING NEW
SOFTWARE TOOLS, but on INTEGRATING them with DATA
and AI TECHNOLOGIES -mostly OPEN SOURCE SOFTWARE.
4.
5. ALEXANDER C. S. HENDORF
MANAGING PARTNER
DATA SCIENCE & AI AT KÖNIGSWEG.
PYTHON SOFTWARE FOUNDATION FELLOW, PYTHON SOFTWAREVERBAND CHAIR,
PYCONDE & PYDATA BERLIN CHAIR, PYDATA FRANKFURT & SÜDWEST, EPS BOARD
@HENDORF python@hendorf.com
6. KÖNIGSWEG ist Gründer lokaler Communities in Frankfurt & Südwest mit je über 1.000 Mitgliedern
– gemeinsam mit Partnern wie HeidelbergCement veranstalten wir Community-Events.
PyData Heidelberg #8: Driving Impact@Scale &
Generative Multimodal Transformer
18. Lighthouse Projects
- Prototypes / Proof of Concepts driven by a
single person on group
- Based on Open Source and Blogposts
lighthouse projects produce results in
reasonable time
- Stakeholders are impressed
19. That looks
great! More of
this!
What about using Python to predict sales
and reports. I build something…
A Stakeholder
A PoC driver
Prototp
ye
The Lighthouse Prototype
20. A Stakeholder
A PoC driver
More ideas
More requirements
The Lighthouse Prototype
More and more features
with various customizations
21.
22. A Stakeholder
A PoC driver
The Prototype
More and more features
with various customizations
Highly motivated, happy to help
Some programming
Some Know-How
Lack of experience in production
Lack of software engineering
Lack of knowledge of inhouse systems
Lack of knowledge in software architecture
Likely not compliant
Potential security and privacy violations
IRRATIC RESULTS- "BUT IT WORKS ON MY COMPUTER"
NO TESTS, NO DOCUMENTATION
NOT SCALABLE
NOT MAINTAINABLE IN TEAM
IMPOSSIBLE TO BRING TO PRODUCTION
23. Mistakes in Lighthouse
Projects
- Highly motivated innovation driver
- Some programming, some know-how
- Believing hypes
Mixed with
- Lack of experience in production
- Lack of software engineering, no tests
- Lack of knowledge of inhouse systems
- Lack of knowledge in software
architecture
- Lack of methodology
24. Requirements to Add Value
Be in production.
- Stability
- Security
- Reliability
- Resilience
25. Best Practise
- Have good ideas
- Try things out!
- Know when to STOP developing further
- Find allies and resources to make it into a real
project
Stakeholders
- Have patience until a solid concept / plan with
requirements is established
- Support with resources and prioritize, enable
26. Measurements
- Build Application based on Prototype idea
completely new based on principles (see
later)
- In-house staff with experienced external
experts (expertise/enabling)
- Transparent process and report to enable
existing team (community factor)
- Involving all stakeholders
-
33. NAVIGATING
THE JOURNEY
LEGACY SYSTEMS ARE HERE TO STAY
INTEGRATION IS REQUIRED
TECHNICAL AND CULTURAL EVOLUTION
OSS SOFTWARE EVOLUTION – SOA CODE
KEEP CURRENT BUSINESS RUNNING
38. 📺
LESSONS LEARNED ABOUT DATA & AI AT
ENTERPRISES AND SMES | PYDATA LONDON
https://www.youtube.com/watch?v=Bp3pUSZ6DpU
5 THINGS YOU WANT TO KNOW ABOUT AI ADOPTION
IN THE ENTERPRISE | PYCONDE & PYDATA BERLIN
https://www.youtube.com/watch?v=0UG_JLUWJOQ&t=77s
LESSONS LEARNED ABOUT DATA & AI AT
ENTERPRISES – DATA CLUB PODCAST
https://www.youtube.com/watch?v=Vms29u9xC3k
39. ALEXANDER C. S. HENDORF
MANAGING PARTNER
DATA SCIENCE & AI AT KÖNIGSWEG.
PYTHON SOFTWARE FOUNDATION FELLOW, PYTHON SOFTWAREVERBAND CHAIR,
PYCONDE & PYDATA BERLIN CHAIR, PYDATA FRANKFURT & SÜDWEST, EPS BOARD
@HENDORF python@hendorf.com
THANK YOU – Q&A