DataTalks.Club

In this podcast episode, we talked with Andrey Cheptsov about ​The future of AI infrastructure.

About the Speaker:

Andrey Cheptsov is the founder and CEO of dstack, an open-source alternative to Kubernetes and Slurm, built to simplify the orchestration of AI infrastructure. Before dstack, Andrey worked at JetBrains for over a decade helping different teams make the best developer tools.

During the event, the guest, Andrey Cheptsov, founder and CEO of dstack, discussed the complexities of AI infrastructure. We explore topics like the challenges of using Kubernetes for AI workloads, the need to rethink container orchestration, and the future of hybrid and cloud-only infrastructures. Andrey also shares insights into the role of on-premise and bare-metal solutions, edge computing, and federated learning.

00:00 Andrey's Career Journey: From JetBrains to DStack

5:00 The Motivation Behind DStack

7:00 Challenges in Machine Learning Infrastructure

10:00 Transitioning from Cloud to On-Prem Solutions

14:30 Reflections on OpenAI's Evolution

17:30 Open Source vs Proprietary Models: A Balanced Perspective

21:01 Monolithic vs. Decentralized AI businesses

22:05 The role of privacy and control in AI for industries like banking and healthcare

30:00 Challenges in training large AI models: GPUs and distributed systems

37:03 DeepSpeed's efficient training approach vs. brute force methods

39:00 Challenges for small and medium businesses: hosting and fine-tuning models

47:01 Managing Kubernetes challenges for AI teams

52:00 Hybrid vs. cloud-only infrastructure

56:03 On-premise vs. bare-metal solutions

58:05 Exploring edge computing and its challenges

🔗 CONNECT WITH ANDREY CHEPTSOV

Twitter -  / andrey_cheptsov  

Linkedin -  / andrey-cheptsov  

GitHub - https://github.com/dstackai/dstack/

Website - https://dstack.ai/

🔗 CONNECT WITH DataTalksClub

Join DataTalks.Club:⁠⁠⁠https://datatalks.club/slack.html⁠⁠⁠

Our events:⁠⁠⁠https://datatalks.club/events.html⁠⁠⁠

Datalike Substack -⁠⁠⁠https://datalike.substack.com/⁠⁠⁠

LinkedIn:⁠⁠⁠  / datatalks-club  ⁠

In this podcast episode, we talked with Tamara Atanasoska about ​building fair AI systems.

About the Speaker:​Tamara works on ML explainability, interpretability and fairness as Open Source Software Engineer at probable. She is a maintainer of fairlearn, contributor to scikit-learn and skops. Tamara has both computer science/ software engineering and a computational linguistics(NLP) background.During the event, the guest discussed their career journey from software engineering to open-source contributions, focusing on explainability in AI through Scikit-learn and Fairlearn. They explored fairness in AI, including challenges in credit loans, hiring, and decision-making, and emphasized the importance of tools, human judgment, and collaboration. The guest also shared their involvement with PyLadies and encouraged contributions to Fairlearn.

00:00 Introduction to the event and the community

01:51 Topic introduction: Linguistic fairness and socio-technical perspectives in AI

02:37 Guest introduction: Tamara’s background and career

03:18 Tamara’s career journey: Software engineering, music tech, and computational linguistics

09:53 Tamara’s background in language and computer science

14:52 Exploring fairness in AI and its impact on society

21:20 Fairness in AI models26:21 Automating fairness analysis in models

32:32 Balancing technical and domain expertise in decision-making

37:13 The role of humans in the loop for fairness

40:02 Joining Probable and working on open-source projects

46:20 Scopes library and its integration with Hugging Face

50:48 PyLadies and community involvement

55:41 The ethos of Scikit-learn and Fairlearn

🔗 CONNECT WITH TAMARA ATANASOSKA

Linkedin - https://www.linkedin.com/in/tamaraatanasoska

GitHub- https://github.com/TamaraAtanasoska

🔗 CONNECT WITH DataTalksClub

Join DataTalks.Club:⁠⁠https://datatalks.club/slack.html⁠⁠

Our events:⁠⁠https://datatalks.club/events.html⁠⁠

Datalike Substack -⁠⁠https://datalike.substack.com/⁠⁠

LinkedIn:⁠⁠  / datatalks-club  

In this podcast episode, we talked with Agita Jaunzeme about Career choices, transitions and promotions in and out of tech.

About the Speaker:

Agita has designed a career spanning DevOps/DataOps engineering, management, community building, education, and facilitation. She has worked on projects across corporate, startup, open source, and non-governmental sectors. Following her passion, she founded an NGO focusing on the inclusion of expats and locals in Porto. Embodying the values of innovation, automation, and continuous learning, Agita provides practical insights on promotions, career pivots, and aligning work with passion and purpose.

During this event, discussed their career journey, starting with their transition from art school to programming and later into DevOps, eventually taking on leadership roles. They explored the challenges of burnout and the importance of volunteering, founding an NGO to support inclusion, gender equality, and sustainability. The conversation also covered key topics like mentorship, the differences between data engineering and data science, and the dynamics of managing volunteers versus employees. Additionally, the guest shared insights on community management, developer relations, and the importance of product vision and team collaboration.

🔗 CONNECT WITH AGITA JAUNZEME

🔗 CONNECT WITH DataTalksClub

In this podcast episode, we talked with Isabella Bicalho about Career advice, learning, and featuring women in ML and AI.

About the Speaker:

Isabella is a Machine Learning Engineer and Data Scientist with three years of hands-on AI development experience. She draws upon her early computational research expertise to develop ML solutions. While contributing to open-source projects, she runs a newsletter dedicated to showcasing women's accomplishments in data science.

During this event, the guest discussed her transition into machine learning, her freelance work in AI, and the growing AI scene in France. She shared insights on freelancing versus full-time work, the value of open-source contributions, and developing both technical and soft skills. The conversation also covered career advice, mentorship, and her Substack series on women in data science, emphasizing leadership, motivation, and career opportunities in tech.

🔗 CONNECT WITH ISABELLA BICALHO

🔗 CONNECT WITH DataTalksClub

Reflection on an Almost Two-Year Journey of Generative AI in Industry – Maria Sukhareva

​About the speaker:

​Maria Sukhareva is a principal key expert in Artificial Intelligence in Siemens with over 15 years of experience at the forefront of generative AI technologies. Known for her keen eye for technological innovation, Maria excels at transforming cutting-edge AI research into practical, value-driven tools that address real-world needs. Her approach is both hands-on and results-focused, with a commitment to creating scalable, long-term solutions that improve communication, streamline complex processes, and empower smarter decision-making. Maria's work reflects a balanced vision, where the power of innovation is met with ethical responsibility, ensuring that her AI projects deliver impactful and production-ready outcomes.

We talked about:

00:00 DataTalks.Club intro

02:13 Career journey: From linguistics to AI

08:02 The Evolution of AI Expertise and its Future

13:10 AI vulnerabilities: Bypassing bot restrictions

17:00 Non-LLM classifiers as a more robust solution

22:56 Risks of chatbot deployment: Reputational and financial

27:13 The role of AI as a tool, not a replacement for human workers

31:41 The role of human translators in the age of AI

34:49 Evolution of English and its Germanic roots

38:44 Beowulf and Old English

39:43 Impact of the Norman occupation on English grammar

42:34 Identifying mushrooms with AI apps and safety precautions

45:08 Decoding ancient languages ​​like Sumerian

49:43 The evolution of machine translation and multilingual models

53:01 Challenges with low-resource languages ​​and inconsistent orthography

57:28 Transition from academia to industry in AI

Join our Slack: https://datatalks.club/slack.html

Our events: https://datatalks.club/events.html

We talked about:

00:00 DataTalks.Club intro

00:00 Large Hadron Collider and Mentorship

02:35 Career overview and transition from physics to data science

07:02 Working at the Large Hadron Collider

09:19 How particles collide and the role of detectors

11:03 Data analysis challenges in particle physics and data science similarities

13:32 Team structure at the Large Hadron Collider

20:05 Explaining the connection between particle physics and data science

23:21 Software engineering practices in particle physics

26:11 Challenges during interviews for data science roles

29:30 Mentoring and offering advice to job seekers

40:03 The STAR method and its value in interviews

50:32 Paid vs unpaid mentorship and finding the right fit

​About the speaker:

​Anastasia is a particle physicist turned data scientist, with experience in large-scale experiments like those at the Large Hadron Collider. She also worked at Blue Yonder, scaling AI-driven solutions for global supply chain giants, and at Kaufland e-commerce, focusing on NLP and search. Anastasia is a mentor for Ml/AI, dedicated to helping her mentees achieve their goals. She is passionate about growing the next generation of data science elite in Germany: from Data Analysts up to ML Engineers.

Join our Slack: https://datatalks .club/slack.html

We talked about:

00:00 DataTalks.Club intro

02:34 Career journey and transition into MLOps

08:41 Dutch agriculture and its challenges

10:36 The concept of "technical debt" in MLOps

13:37 Trade-offs in MLOps: moving fast vs. doing things right

14:05 Building teams and the role of coordination in MLOps

16:58 Key roles in an MLOps team: evangelists and tech translators

23:01 Role of the MLOps team in an organization

25:19 How MLOps teams assist product teams

27 :56 Standardizing practices in MLOps

32:46 Getting feedback and creating buy-in from data scientists

36:55 The importance of addressing pain points in MLOps

39:06 Best practices and tools for standardizing MLOps processes

42:31 Value of data versioning and reproducibility

44:22 When to start thinking about data versioning

45:10 Importance of data science experience for MLOps

46:06 Skill mix needed in MLOps teams

47:33 Building a diverse MLOps team

48:18 Best practices for implementing MLOps in new teams

49:52 Starting with CI/CD in MLOps

51:21 Key components for a complete MLOps setup

53:08 Role of package registries in MLOps

54:12 Using Docker vs. packages in MLOps

57:56 Examples of MLOps success and failure stories

1:00:54 What MLOps is in simple terms

1:01:58 The complexity of achieving easy deployment, monitoring, and maintenance

Join our Slack: https://datatalks .club/slack.html

We talked about:

00:00 DataTalks.Club intro

About the speaker:

Rachel is an urban data scientist dedicated to creating liveable cities through the innovative use of data. With a background in geography, and a masters in urban data science, she blends qualitative and quantitative analysis to tackle urban challenges. Her aim is to integrate data driven techniques with urban design to foster sustainable and equitable urban environments. 

Links: - https://datamall.lta.gov.sg/content/datamall/en/dynamic-data.html

We talked about:

00:00 DataTalks.Club intro

00:00 DataTalks.Club anniversary "Ask Me Anything" event with Alexey Grigorev

02:29 The founding of DataTalks .Club

03:52 Alexey's transition from Java work to DataTalks.Club

04:58 Growth and success of DataTalks.Club courses

12:04 Motivation behind creating a free-to-learn community

24:03 Staying updated in data science through pet projects

26 :37 Hosting a second podcast and maintaining programming skills

28:56 Skepticism about LLMs and their relevance

31:53 Transitioning to DataTalks.Club and personal reflections

33:32 Memorable moments and the first event's success

36:19 Community building during the pandemic

38:31 AI's impact on data analysts and future roles

42:24 Discussion on AI in healthcare

44:37 Age and reflections on personal milestones

47:54 Building communities and personal connections

49:34 Future goals for the community and courses

51:18 Community involvement and engagement strategies

53:46 Ideas for competitions and hackathons

54:20 Inviting guests to the podcast

55:29 Course updates and future workshops

56:27 Podcast preparation and research process

58:30 Career opportunities in data science and transitioning fields

1:01 :10 Book recommendations and personal reading experiences

About the speaker:

Alexey Grigorev is the founder of DataTalks.Club.

Join our slack: https://datatalks.club/slack.html

We talked about:

00:00 DataTalks.Club intro

08:06 Background and career journey of Katarzyna

09:06 Transition from linguistics to computational linguistics

11:38 Merging linguistics and computer science

15:25 Understanding phonetics and morpho-syntax

17:28 Exploring morpho-syntax and its relation to grammar

20:33 Connection between phonetics and speech disorders

24:41 Improvement of voice recognition systems

27:31 Overview of speech recognition technology

30:24 Challenges of ASR systems with atypical speech

30:53 Strategies for improving recognition of disordered speech

37:07 Data augmentation for training models

40:17 Transfer learning in speech recognition

42:18 Challenges of collecting data for various speech disorders

44:31 Stammering and its connection to fluency issues

45:16 Polish consonant combinations and pronunciation challenges

46:17 Use of Amazon Transcribe for generating podcast transcripts

47:28 Role of language models in speech recognition

49:19 Contextual understanding in speech recognition

51:27 How voice recognition systems analyze utterances

54:05 Personalization of ASR models for individuals

56:25 Language disorders and their impact on communication

58:00 Applications of speech recognition technology

1:00:34 Challenges of personalized and universal models

1:01:23 Voice recognition in automotive applications

1:03:27 Humorous voice recognition failures in cars

1:04:13 Closing remarks and reflections on the discussion

About the speaker:

Katarzyna is a computational linguist with over 10 years of experience in NLP and speech recognition. She has developed language models for automotive brands like Audi and Porsche and specializes in phonetics, morpho-syntax, and sentiment analysis.

Kasia also teaches at the University of Warsaw and is passionate about human-centered AI and multilingual NLP.

Join our slack: https://datatalks.club/slack.html