Episodes

  • #26 AI Trends in 2025 - DODS
    Jan 12 2025

    ๐–๐ก๐š๐ญ ๐€๐ˆ ๐ญ๐ซ๐ž๐ง๐๐ฌ ๐ฐ๐ข๐ฅ๐ฅ ๐๐จ๐ฆ๐ข๐ง๐š๐ญ๐ž ๐ข๐ง 2025? ๐‡๐จ๐ฐ ๐ฐ๐ข๐ฅ๐ฅ ๐›๐ฎ๐ฌ๐ข๐ง๐ž๐ฌ๐ฌ๐ž๐ฌ ๐š๐๐š๐ฉ๐ญ? ๐€๐ง๐ ๐ฐ๐ก๐š๐ญ ๐๐จ๐ž๐ฌ ๐ข๐ญ ๐š๐ฅ๐ฅ ๐ฆ๐ž๐š๐ง ๐Ÿ๐จ๐ซ ๐ฉ๐ซ๐จ๐Ÿ๐ž๐ฌ๐ฌ๐ข๐จ๐ง๐š๐ฅ๐ฌ ๐ฅ๐ข๐ค๐ž ๐ฒ๐จ๐ฎ?

    In our new 26th episode, โ€œ๐€๐ˆ ๐“๐ซ๐ž๐ง๐๐ฌ ๐ข๐ง 2025โ€, we sit down with Jasmin/Kate to discuss the key trends and changes shaping the future of AI and the data field:

    • AI Trends for 2025 for business
    • The shifts coming to hiring processes for data roles
    • What professionals can expect in this evolving industry


    ๐ŸŽ™ New episode drops today! Listen & Subscribe


    โœจ Interesting facts for you:

    • Nearly 60% of businesses plan to increase their AI investments in 2025 (Hosting Survey).
    • Big Data Specialist is the fastest-growing job, with Data Scientists and Analysts ranked 11th (World Economic Forum).

    If you enjoy our podcast, please consider:

    โญ Leaving a review on Spotify / Apple Podcasts

    ๐Ÿ“ฉ Subscribe

    ๐Ÿ’ฌ Leave a comment or ๐Ÿ‘ to support us!

    Your engagement means the world to us.โฃ๏ธ

    #DataScience #DODS #DiariesOfADataScientist #AICommunity

    ๐Ÿชฝ Follow Jasmin on LinkedIn: https://www.linkedin.com/in/jasmin-weimueller-bsc2018/

    ๐Ÿชฝ Follow Kate on LinkedIn: https://www.linkedin.com/in/kate-nazarova-data-science/

    ๐Ÿชฝ Subscribe to our official DODS page: https://www.linkedin.com/company/diaries-of-data-scientist/

    Follow us on Medium๐Ÿ‘‡

    ๐Ÿ–‡ Jasminโ€™s Medium page: https://medium.com/@JasminWhy

    ๐Ÿ–‡ Kateโ€™s Medium page: https://medium.com/@Kate_in_DS


    Join us on other platforms:

    ๐ŸŽง Apple: https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://podcasts.apple.com/us/podcast/diaries-of-a-data-scientist/id1710961657&ved=2ahUKEwjhm8PdhMWIAxV6qZUCHYZsCDwQFnoECBsQAQ&usg=AOvVaw1deaPC2MF6aWM69-SKSRoH

    ๐ŸŽง Amazon: https://amzn.asia/d/7J3UkTE

    ๐ŸŽง Podimo: https://podimo.com/de/shows/diaries-of-a-data-scientist

    ๐ŸŽง Podscribe: https://app.podscribe.ai/series/2353052

    Show More Show Less
    36 mins
  • #25 Thinking Clearly - a Data Science Guide - DODS
    Dec 22 2024

    #Christmas Special๐ŸŽ„๐˜๐จ๐ฎ๐ซ ๐ƒ๐š๐ญ๐š ๐ข๐ฌ ๐‡๐จ๐ง๐ž๐ฌ๐ญ. ๐˜๐จ๐ฎ๐ซ ๐๐ข๐š๐ฌ ๐ข๐ฌ ๐๐จ๐ญ.

    As 2024 comes to a close, it's time to pause and reflect on how we make decisions in data-driven projects. In this Christmas Special, weโ€™re diving into the fascinating world of โ€๐‚๐จ๐ ๐ง๐ข๐ญ๐ข๐ฏ๐ž ๐๐ข๐š๐ฌ๐ž๐ฌ ๐ข๐ง ๐ƒ๐š๐ญ๐š ๐’๐œ๐ข๐ž๐ง๐œ๐žโ€. ๐Ÿง โœจ

    In this episode, you'll learnโ€ฆ

    ๐ŸŽฏ See how your own bias might be sabotaging your data science thinking process

    ๐ŸŽฏ Reflect on the decision-making process in 2024

    ๐ŸŽฏ Explore strategies to refine and elevate your approaches for 2025

    In fact, ๐Ÿ‘พ massive algorithms built for mass usage cannot be created by one person aloneโ€”and cognitive biases are one of the reasons why.


    ๐˜๐จ๐ฎ๐ซ ๐๐š๐ญ๐š ๐ฆ๐ข๐ ๐ก๐ญ ๐ง๐จ๐ญ ๐›๐ž ๐ฅ๐ฒ๐ข๐ง๐ โ€”๐›๐ฎ๐ญ ๐ฒ๐จ๐ฎ๐ซ ๐›๐ข๐š๐ฌ ๐ฆ๐ข๐ ๐ก๐ญ ๐›๐ž.


    If you like our podcast, please consider leaving a review on Spotify, subscribe on Youtube and leave a comment and/or ๐Ÿ‘! This would help us a lot ๐Ÿฉต

    #datascience #dods #diariesofadatascientist #aicommunity


    Reading Recommendations:

    • Thinking Fast and Slow - Daniel Kahneman
    • The art of thinking clearly - Rolf Dobelli
    • Quit - Annie Duke


    ๐Ÿชฝ Follow Jasmin on LinkedIn: https://www.linkedin.com/in/jasmin-weimueller-bsc2018/

    ๐Ÿชฝ Follow Kate on LinkedIn: https://www.linkedin.com/in/kate-nazarova-data-science/

    ๐Ÿชฝ Subscribe to our official DODS page: https://www.linkedin.com/company/diaries-of-data-scientist/

    Follow us on Medium๐Ÿ‘‡

    ๐Ÿ–‡ Jasminโ€™s Medium page: https://medium.com/@JasminWhy

    ๐Ÿ–‡ Kateโ€™s Medium page: https://medium.com/@Kate_in_DS

    Join us on other platforms:

    ๐ŸŽง Spotify: https://open.spotify.com/show/1DAelRe22W8vBHK7rTU361?si=4e4f3d7bc67546cc

    ๐ŸŽง Apple: https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://podcasts.apple.com/us/podcast/diaries-of-a-data-scientist/id1710961657&ved=2ahUKEwjhm8PdhMWIAxV6qZUCHYZsCDwQFnoECBsQAQ&usg=AOvVaw1deaPC2MF6aWM69-SKSRoH

    ๐ŸŽง Amazon: https://amzn.asia/d/7J3UkTE

    ๐ŸŽง Podimo: https://podimo.com/de/shows/diaries-of-a-data-scientist

    ๐ŸŽง Podscribe: https://app.podscribe.ai/series/2353052

    Show More Show Less
    42 mins
  • #24 Mentoring Insights: Top 5 Data Science Questions from Singapore - DODS
    Nov 24 2024

    Ever found yourself wondering, โ€œIs data science really for me?โ€ Or maybe youโ€™ve asked, โ€œWhat does it actually take to become a data scientist?โ€ Weโ€™ve been there, too!

    Thatโ€™s why in our latest episode โ€œMentoring Insights: Top 5 Data Science Questions from Singaporeโ€, we dove into some of the most common (and often intimidating) questions about breaking into the field of data science:

    ๐ŸŒŸ How do I know if data science is for me?

    ๐ŸŽ“ Which qualifications do I really need?

    ๐Ÿ’ก Why isnโ€™t a single course or bootcamp enough?


    If youโ€™re considering a career in data science or trying to navigate the next step, this episode is for you. ๐Ÿ’™ And remember, itโ€™s okay to feel overwhelmed at first; every expert you admire was once a beginner!


    If you like our podcast, please consider leaving a review on Spotify, subscribe on Youtube and leave a comment and/or ๐Ÿ‘! This would help us a lot โ™ฅ๏ธ

    #datascience #dods #diariesofadatascientist #aicommunity


    ๐Ÿชฝ Follow Jasmin on LinkedIn: https://www.linkedin.com/in/jasmin-weimueller-bsc2018/

    ๐Ÿชฝ Follow Kate on LinkedIn: https://www.linkedin.com/in/kate-nazarova-data-science/

    ๐Ÿชฝ Subscribe to our official DODS page: https://www.linkedin.com/company/diaries-of-data-scientist/

    Follow us on Medium๐Ÿ‘‡

    ๐Ÿ–‡ Jasminโ€™s Medium page: https://medium.com/@JasminWhy

    ๐Ÿ–‡ Kateโ€™s Medium page: https://medium.com/@Kate_in_DS

    Join us on other platforms:

    ๐ŸŽง Spotify: https://open.spotify.com/show/1DAelRe22W8vBHK7rTU361?si=4e4f3d7bc67546cc

    ๐ŸŽง Apple: https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://podcasts.apple.com/us/podcast/diaries-of-a-data-scientist/id1710961657&ved=2ahUKEwjhm8PdhMWIAxV6qZUCHYZsCDwQFnoECBsQAQ&usg=AOvVaw1deaPC2MF6aWM69-SKSRoH

    ๐ŸŽง Amazon: https://amzn.asia/d/7J3UkTE

    ๐ŸŽง Podimo: https://podimo.com/de/shows/diaries-of-a-data-scientist

    ๐ŸŽง Podscribe: https://app.podscribe.ai/series/2353052

    Show More Show Less
    50 mins
  • #23 In AI we trust ?!
    Oct 27 2024

    In our latest episode of Diaries of a Data Scientist we dive into one of the most pressing challenges we face today: ๐‡๐จ๐ฐ ๐๐จ ๐ฐ๐ž ๐›๐ฎ๐ข๐ฅ๐ ๐ญ๐ซ๐ฎ๐ฌ๐ญ ๐ข๐ง ๐€๐ˆ ๐š๐ง๐ ๐ƒ๐š๐ญ๐š ๐’๐œ๐ข๐ž๐ง๐œ๐ž ๐ฌ๐จ๐ฅ๐ฎ๐ญ๐ข๐จ๐ง๐ฌ? ๐Ÿ‘€๐Ÿ‘€

    Kate recently met with a team of data scientists in Brazil, where key questions many of us have faced were raised:

    โ€” How do we make business teams trust our modelsโ“

    โ€” How can we translate complex technical concepts into something business people ๐ซ๐ž๐š๐ฅ๐ฅ๐ฒ understandโ“

    โ€” And how do we ensure our models deliver exactly whatโ€™s neededโ“

    In this episode, we break down ๐ฌ๐ญ๐ซ๐š๐ญ๐ž๐ ๐ข๐ž๐ฌ you can use to bridge the gap between tech and business, build transparency, and most importantlyโ€”gain trust.

    Plus, weโ€™re sharing ๐š๐œ๐ญ๐ข๐จ๐ง๐š๐›๐ฅ๐ž ๐ ๐ฎ๐ข๐๐š๐ง๐œ๐ž on making your strategies more accessible and trusted, drawing insights from our experience, the Technology Acceptance Model (TAM), and how explainability is shaping trust in tech.

    Trust me, you donโ€™t want to miss this one. ๐Ÿ˜‰

    If you like our podcast, please consider leaving a review on Spotify, subscribe on Youtube and leave a comment and/or ๐Ÿ‘! This would help us a lot โ™ฅ๏ธ

    #datascience #dods #diariesofadatascientist #aicommunity

    ๐Ÿชฝ Follow Jasmin on LinkedIn: https://www.linkedin.com/in/jasmin-weimueller-bsc2018/

    ๐Ÿชฝ Follow Kate on LinkedIn: https://www.linkedin.com/in/kate-nazarova-data-science/

    ๐Ÿชฝ Subscribe to our official DODS page: https://www.linkedin.com/company/diaries-of-data-scientist/

    Follow us on Medium๐Ÿ‘‡

    ๐Ÿ–‡ Jasminโ€™s Medium page: https://medium.com/@JasminWhy

    ๐Ÿ–‡ Kateโ€™s Medium page: https://medium.com/@Kate_in_DS

    Join us on other platforms:

    ๐ŸŽง Spotify: https://open.spotify.com/show/1DAelRe22W8vBHK7rTU361?si=4e4f3d7bc67546cc

    ๐ŸŽง Apple: https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://podcasts.apple.com/us/podcast/diaries-of-a-data-scientist/id1710961657&ved=2ahUKEwjhm8PdhMWIAxV6qZUCHYZsCDwQFnoECBsQAQ&usg=AOvVaw1deaPC2MF6aWM69-SKSRoH

    ๐ŸŽง Amazon: https://amzn.asia/d/7J3UkTE

    ๐ŸŽง Podimo: https://podimo.com/de/shows/diaries-of-a-data-scientist

    ๐ŸŽง Podscribe: https://app.podscribe.ai/series/2353052

    Show More Show Less
    1 hr and 11 mins
  • #22 David vs. Goliath: Open Source Takes on Generative AI Giants - DODS
    Oct 6 2024
    ๐–๐ž๐ฅ๐œ๐จ๐ฆ๐ž ๐›๐š๐œ๐ค ๐ญ๐จ ๐ญ๐ก๐ž ๐„๐ฉ๐ข๐ฌ๐จ๐๐ž 21! ๐ŸŽ™ Have you ever wondered how much control you truly have over your Gen. AI models? What about the protection of your data? ๐Ÿค” E๐ฉ๐ข๐ฌ๐จ๐๐ž #22 ๐จ๐Ÿ โ€œ๐ƒ๐ข๐š๐ซ๐ข๐ž๐ฌ ๐จ๐Ÿ ๐š ๐ƒ๐š๐ญ๐š ๐’๐œ๐ข๐ž๐ง๐ญ๐ข๐ฌ๐ญโ€ explores a David vs. Goliath like story: Open source taking on Generative AI Giants. ? Why should you care about models you can run ๐ฅ๐จ๐œ๐š๐ฅ๐ฅ๐ฒ or in your own ๐ฉ๐ซ๐ข๐ฏ๐š๐ญ๐ž ๐œ๐ฅ๐จ๐ฎ๐ ? What if you could avoid paying a surplus on each token processed ? And how valuable is the global ๐œ๐จ๐ฆ๐ฆ๐ฎ๐ง๐ข๐ญ๐ฒ constantly improving and innovating these models? IWe also cover relevant ๐จ๐ฉ๐ž๐ง-๐ฌ๐จ๐ฎ๐ซ๐œ๐ž ๐๐š๐ญ๐š๐ฌ๐ž๐ญ๐ฌ and models for ๐ญ๐ž๐ฑ๐ญ-๐ญ๐จ-๐ญ๐ž๐ฑ๐ญ and ๐ญ๐ž๐ฑ๐ญ-๐ญ๐จ-๐ข๐ฆ๐š๐ ๐ž generation. If you're ready to extend your horizon beyond the standard Gen AI providers, this episode is for you! ๐Ÿชฝ Follow Jasmin on LinkedIn: https://www.linkedin.com/in/jasmin-weimueller-bsc2018/ ๐Ÿชฝ Follow Kate on LinkedIn: https://www.linkedin.com/in/kate-nazarova-data-science/ ๐Ÿชฝ Subscribe to our official DODS page: https://www.linkedin.com/company/diaries-of-data-scientist/ Follow us on Medium๐Ÿ‘‡ ๐Ÿ–‡ Jasminโ€™s Medium page: https://medium.com/@JasminWhy ๐Ÿ–‡ Kateโ€™s Medium page: https://medium.com/@Kate_in_DS Join us on other platforms: ๐ŸŽง Spotify: https://open.spotify.com/show/1DAelRe22W8vBHK7rTU361?si=4e4f3d7bc67546cc ๐ŸŽง Apple: https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://podcasts.apple.com/us/podcast/diaries-of-a-data-scientist/id1710961657&ved=2ahUKEwjhm8PdhMWIAxV6qZUCHYZsCDwQFnoECBsQAQ&usg=AOvVaw1deaPC2MF6aWM69-SKSRoH ๐ŸŽง Amazon: https://amzn.asia/d/7J3UkTE ๐ŸŽง Podimo: https://podimo.com/de/shows/diaries-of-a-data-scientist ๐ŸŽง Podscribe: https://app.podscribe.ai/series/2353052 Useful links & Resources: State of Open Source AI :https://github.blog/news-insights/research/the-state-of-open-source-and-ai/ LLaMA https://about.fb.com/news/2024/07/open-source-ai-is-the-path-forward/ **LAION-5B: https://huggingface.co/datasets/danielz01/laion-5b** **The Pile: https://huggingface.co/datasets/EleutherAI/pile** **C4: https://huggingface.co/datasets/legacy-datasets/c4** GPT-Neo / GPT-J: https://huggingface.co/docs/transformers/en/model_doc/gpt_neo; https://huggingface.co/docs/transformers/en/model_doc/gptj **Mixtral 8x7B: https://huggingface.co/mistralai/Mixtral-8x7B-v0.1** **BLOOM: https://bigscience.huggingface.co/blog/bloom** **T5: https://huggingface.co/docs/transformers/en/model_doc/t5** **LLaMA: https://ai.meta.com/blog/llama-3-2-connect-2024-vision-edge-mobile-devices/** Stable Diffusion: https://huggingface.co/models?other=stable-diffusion **DALL-E Mini: https://wandb.ai/dalle-mini/dalle-mini/reports/DALL-E-Mini-Explained--Vmlldzo4NjIxODA; https://github.com/borisdayma/dalle-mini** **FLUX.1: https://www.bentoml.com/blog/a-guide-to-open-source-image-generation-models**
    Show More Show Less
    40 mins
  • #21 Life Update: Move to Brazil - DODS
    Sep 19 2024

    ๐–๐ž๐ฅ๐œ๐จ๐ฆ๐ž ๐›๐š๐œ๐ค ๐ญ๐จ ๐ญ๐ก๐ž ๐„๐ฉ๐ข๐ฌ๐จ๐๐ž 21! ๐ŸŽ™

    Today, weโ€™re doing something a little different. We thought weโ€™d take a break from the usual deep dives into data science and just chat with you about ๐—…๐—‚๐–ฟ๐–พ ๐—Ž๐—‰๐–ฝ๐–บ๐—๐–พ๐—Œ, ๐—‡๐–พ๐— ๐—‚๐–ฝ๐–พ๐–บ๐—Œ weโ€™re cooking up for the next few months, sharing ๐—Œ๐—ˆ๐—†๐–พ ๐–ฟ๐—Ž๐—‡ ๐—Œ๐—๐—ˆ๐—‹๐—‚๐–พ๐—Œ ๐–บ๐—‡๐–ฝ ๐–ผ๐—ˆ๐—†๐—‰๐–บ๐—‹๐—‚๐—Œ๐—ˆ๐—‡๐—Œ ๐–ป๐–พ๐—๐—๐–พ๐–พ๐—‡ ๐—…๐–บ๐—‡๐—€๐—Ž๐–บ๐—€๐–พ๐—Œ ๐–บ๐—‡๐–ฝ ๐–ผ๐—Ž๐—…๐—๐—Ž๐—‹๐–พ๐—Œ.

    Let us know if you enjoy this type of content! Weโ€™d be happy to sprinkle in more personal stories alongside the tech discussions.


    If you like our podcast, please consider leaving a review on Spotify, subscribe on Youtube and leave a comment and/or ๐Ÿ‘! This would help us a lot โ™ฅ๏ธ


    #datascience #dods #diariesofadatascientist #aicommunity

    Show More Show Less
    13 mins
  • # 20 Who benefits from Open Source AI?
    Sep 15 2024

    ๐™ธ๐—‡โ€€๐–บโ€€๐—๐—ˆ๐—ˆ๐–ฝ๐—Œโ€€๐—ˆ๐–ฟโ€€๐—๐—๐–พโ€€๐–ฝ๐–บ๐—๐–บโ€€๐—Œ๐–ผ๐—‚๐–พ๐—‡๐–ผ๐–พโ€€๐—๐—ˆ๐—‹๐—…๐–ฝ,โ€€๐—ˆ๐—‡๐–พโ€€๐–ผ๐–บ๐—‡โ€€๐—Œ๐—Ž๐—‹๐–พ๐—…๐—’โ€€๐–ฟ๐—‚๐—‡๐–ฝโ€€๐—๐—๐–พ๐—‚๐—‹โ€€๐—Œ๐—‰๐—ˆ๐—! ๐Ÿ‘พ

    Whether you're a developer or an entrepreneur, open-source communities ๐Ÿ‘ฅ are a game-changer. This #๐Ÿค๐Ÿข๐—๐— ๐–พ๐—‰๐—‚๐—Œ๐—ˆ๐–ฝ๐–พ ๐—ˆ๐–ฟ ๐–ฃ๐–บ๐—‚๐—‹๐—‚๐–พ๐—Œ ๐—ˆ๐–ฟ ๐–บ ๐–ฃ๐–บ๐—๐–บ ๐–ฒ๐–ผ๐—‚๐–พ๐—‡๐—๐—‚๐—Œ๐— is ๐–ธ๐—ˆ๐—Ž๐—‹ ๐–ฆ๐—Ž๐—‚๐–ฝ๐–พ to not just improving your skills but also:

    ๐Ÿ†’ earning trust

    ๐Ÿ†’ credibility

    ๐Ÿ†’ and making a real impact in the tech ecosystem

    ๐ŸŽ™๏ธ Join @Jasmin/@Kate and Kate/Jasmin as we dive deep into the advantages of contributing to open-source packages, from boosting your career to building valuable connections in both tech and business.


    If you like our podcast, please consider leaving a review on Spotify, subscribe on Youtube and leave a comment and/or ๐Ÿ‘! This would help us a lot โ™ฅ๏ธ

    #datascience #dods #diariesofadatascientist #aicommunity

    ๐Ÿ‘‰ P.S. Feel free to check out Episode 15 if you're interested in learning more about interdisciplinary positions with a strong focus on open-source communities

    Show More Show Less
    49 mins
  • #19 Generative AI in 2024: Key Milestones You Need to Know - DODS
    Aug 18 2024

    In this episode, we dive into the most significant milestones in generative AI during 2024 and explore how these advancements are transforming industries worldwide.

    Whether you're a tech enthusiast, industry professional, or just curious about the future of AI, this episode provides a comprehensive overview of how generative AI is making its mark in 2024 and beyond.

    Some links to further check out resources:

    https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai

    https://openai.com/index/hello-gpt-4o/

    https://www.ibm.com/think/insights/artificial-intelligence-trends

    https://www.etoro.com/news-and-analysis/press-releases/etoros-first-ai-generated-ad-campaign-to-be-aired-during-paris-olympics/

    If you like our podcast, please consider leaving a review on Spotify, subscribe on Youtube and leave a comment and/or ๐Ÿ‘! This would help us a lot โ™ฅ๏ธ

    #datascience #dods #diariesofadatascientist #aicommunity

    Below we attach a list of models that is helpful to catch up with the most relevant ones from 2024 so far:

    • Llama 3.1: Meta's latest version in the Llama series, offering significant improvements in both language understanding and generation capabilities. https://llama.meta.com
    • GPT-4o and GPT-4o mini: Enhanced versions of OpenAI's GPT series, focusing on more efficient and contextually aware text generation.
    • Phi 3: A new model developed for complex multimodal tasks, integrating text, audio, and visual data processing. (SLM): https://azure.microsoft.com/en-us/blog/introducing-phi-3-redefining-whats-possible-with-slms/
    • Mistral Large 2: Known for its low latency and high performance, particularly in real-time applications (The GitHub Blog) (ThinkPalm).
    • DeepDream: An updated version of Google's model that generates visually striking and abstract images using deep neural networks (ThinkPalm).
    • GitHub Copilot: A collaborative tool that assists developers by providing context-aware code completions, now enhanced with broader language support and improved interactive documentation suggestions (ThinkPalm) (The GitHub Blog).
    • Microsoft 365 CoPilot, Studio
    • Complexity of Generative AI Adoption in Corporate

    Recent Generative AI Models Developed (January 2024 - July 2024)

    Show More Show Less
    45 mins