The Virtual Feature Store 🔮 Featureform allows data scientists to define, manage, and serve machine learning features across their organization.
SF
- Tweets 462
- Following 28
- Followers 334
- Likes 248
🌟 We had an awesome happy hour last week!
Big shoutout to the @modal_labs team, including @alecpowell and @charles_irl, for co-sponsoring last week's happy hour with us. We had a blast with you guys! 🚀
6
359
🚂 Thinking of hopping on the AI hype train?
Before you do, listen to our latest podcast, where Gautam Krishnamurthi of GreatPoint Ventures chats with Featureform CEO Simba Khadder about the rapidly evolving world of AI and its impacts on the future of venture capital investing.
1
287
Visit our booth at the Databricks Data + AI Summit this week (we're booth #6, near the pickleball court!), where you can learn more about our open source and enterprise products, get a live demo and, most importantly, get some awesome swag including a Feature Creature plushie!
157
🚂 Should all MLOps companies jump on the LLM hype train?
Join Simba Khadder and Union.ai's founder, Ketan Umare, as they tackle this topic and more on our latest episode of the MLOps Weekly Podcast!
📺 Watch the full podcast here! buff.ly/3xgXBKV
491
Struggling to get your ML models out of your notebook and into production? MLOps to the rescue! 🚀
1
1
81
🕳️ The Pitfall of Discrepant Feature Definitions in ML Systems 🕳️
🔍 The Issue at Hand
Feature engineering is the backbone of any machine learning model. It's where raw data is transformed into predictive signals to feed into a model.
1
1
88
Tons of ML teams are stuck under mountains of technical and documentation debt. It might feel like you need a degree in library science to manage it all! 🤓
There's a pathway through this dense forest of information 🌟 Featureform and MLFlow 🌟
2
75
🤔 Ever look at a training set or model feature and have no idea how it was created? You're not alone! Sometimes documentation takes a back seat. But hey, we've all been there, right? 😅
1
1
45
🔥 Deploying features without drift monitoring? That's like leaving for vacation with your stove on — and nobody likes coming back to extra crispy features!
1
1
48
📦 Feature stores ARE NOT just storage for features! Did you know there are three distinct Feature Store architectures? 🤔
🧵👇
1
3
119
Check the full episode here: buff.ly/3QJkcp5
The #MLOps Weekly Podcast is available on Spotify, Apple, and Google podcasts!
37
🎙 EPISODE DROP: This week on the #MLOps Weekly Podcast, @simba_khadder chats w/ @mlukichev , Co-Founder and CTO at @telmai1 🎉
Check out Max's thoughts on why Data Management is such a difficult problem to tackle in the clip below 👇
1
1
1
161
🤔 Attending @_odsc West today?
📅Join us this afternoon for "Feature Stores in Practice: Build & Deploy a Model with Featureform, Redis, Databricks, and Amazon SageMaker" a hands-on workshop hosted by our CEO, @simba_khadder! 🎉
🔗 Register here: buff.ly/2NHxV0N
2
194
🔈 Louder for those in the back: FEATURE STORES ARE NOT ABOUT STORAGE👏 💅
Tune into the Feature Store Summit on October 11th for an insightful talk on the different types of #featurestore architectures
🔗 Register here: buff.ly/3LT2vlw
1
63
🚀🎉 #Hacktoberfest with Featureform is here! 🎉🚀
We're thrilled to invite you to join us this Hacktoberfest & celebrate the wonder of open-source.
🎁 Every merged contribution gets you exclusive Featureform swag & a chance to win prizes! 💰🎉
🔗:buff.ly/48JdGXD
1
113
🔥 New at Featureform: Quickstart Notebook with the On-Demand Features API!
For those dealing with real-time data challenges: Our On-Demand Features API lets you compute features at request time, eliminating race conditions 🎉
🔗 Check it out here: buff.ly/48CHKnX
2
102
💼 Business: Update the model plz.
🧑🔬 Data Scientist: How hard can it be?
#datascience #dataops #ml #mlengineer #mlops #dataengineering
1
55
🤔 How important is team maturity and a phased approach when evaluating #MLOps tools?
@mlopscommunity Founder, @Dpbrinkm shares his thoughts in the clip below 👇
Check out his full episode on the MLOps Weekly Podcast here: buff.ly/3PtZr0f
2
46
🎇 When the project is just piles of pyspark functions with no comments or readme in sight
#datascience #mlops #mlengineer #dataops #dataengineering #pyspark
1
2
109
🔥 Join us for “Serving Batch, Streaming, and Real-time Features with Featureform” on Thurs, Sept 21st at 8 AM PT🔥
Learn how Featureform’s Python API can be easily integrated into your data science workflow! 🙌
🎟 Sign-up here: buff.ly/3sYgRKE
2
122
📈📉 Correlation isn't causation and ML has a funny way of making you realize you don't understand the business as well as you think you do.
#datascience #mlengineering #mlops #dataops #dataengineering #ml #interpretability
1
1
84
Dive into the details in our blog post and see how Featureform can help you unlock the value of data for machine learning in a reproducible, collaborative manner.
📂 Check out the repo: buff.ly/45MGdKx
🔗 Link to blog post: bit.ly/3Kc3yMr
25
🔹 From Experimentation To Production: During the development and experimentation stage, use Featureform’s auto-variants capabilities to jump straight into the hard work of analyzing, understanding, and generating features from your data sources.
1
28
🔹 Organizing Feature Development With Named Runs: By using auto-generated variants and setting session run names once, you’re less prone to mistyping or having to juggle names across notebook cells.
1
15
🔹 Auto-Generated Variants: Featureform’s auto-generated variants capability allows data science practitioners to jump straight into developing and evaluating feature logic while experiencing the benefits of versioning.
1
16
🔹 Tracking Versions with Variants: Using variants, you can easily version your transformations and feature definitions. Featureform supports versioning for all resources including providers, sources, features and labels, entities, training sets, and models.
1
15
🔹Variants vs Versions: Variants depict different forms of something coexisting, aligning more with the reality in #ml where different dataset transformations or feature combinations aren't sequential versions, but parallel alternatives under simultaneous test & trial.
1
22
🔍 Dive into the world of feature versioning with Featureform! 🔍
We explore how Featureform supports versioning with "variants" - a term that more accurately reflects the rich diversity and non-linear progression of changes typical in the machine learning world.
🧵
1
82
☕️ Nothing like being woken up early Sunday morning by slack notifications because Airflow was having issues 😴🛏️
#datascience #dataops #mlops #mlengineer #dataengineer #airflow #featurestores
54
💢😧: "I don't understand, what do you mean the features in our data warehouse aren't available for model serving & inference in prod?"
#datascience #mlops #dataengineering #mlengineer #ml #dataops #featurestores
1
1
189
😭 "What do you mean we use a different framework for serving features?"
#datascience #mlops #mlengineer #dataops #dataengineer #ml
1
229
💼 "We're a customer centric and data driven company with a strong engineering practice...
that doesn't really have mature practices in place for ML". 👀
#DataScience #mlops #dataengineering #dataops #ml #datascientists
3
139
🎇❄️🧱🐍📓 Always fun trying to retraining on spaghetti code
#datascience #mlops #dataengineering #datascience #machinelearning
5
144
🌐 Featureform is dedicated to helping teams build and scale production ML. Our latest piece tackles the intricate task of measuring ML platforms. #mlplatforms
1
7
🛣️ Use our framework to strategize and plan your roadmap to ML Platform Adoption Readiness. Dive into the full blog post for more insights. bit.ly/3OOagLq #dataops
1
5
📌 Key Takeaways:
✔️ ML Platforms should be treated "as-a-Product".
✔️ The maturity of the MLOps stack is deeply tied to North Star metrics.
✔️ Platform Adoption is a step above ML Product Delivery. #mlengineer
1
8
🚀 The democratization of access to data science tools has transformed industries, emphasizing the importance of MLOps in today's competitive landscape. #mlops
1
4
🌟 The rise of data science and ML tools has ushered in a new era where the fastest and most reliable ideas dominate the market. #datascience
1
1
66
🔬 Probably should have made sure your feature engineering pipeline mimicked the one you wrote in dev & accounted for the edge cases🤔
#mlops #dataops #datascience #mlengineer #featurestores #dataengineering
1
4
126
Finally about us some more det's about us, the sponsors for this @mlopscommunity #womeninmlops event 👇
#mlops #datascientists #mlengineers #dataops #dataengineering #mlopscommunity #featurestores #llmops
1
2
85
Big thanks to the @mlopscommunity crew for making the event happen!
@RahulParundekar
@anshuman_01_ (@ChatNBX @NimbleBoxAI)
@Afrootiee (@goafterwork)
@adamboazbecker
1
1
4
546
Big thanks to our panelists for contributing their thoughts, experiences, & lessons around:
- solo-founding
- the YC route
- career crafting, esp, in ML
- balancing leadership & technical now-how
@DataCereal (@VertaAI)
@biankaroy_ (@NVIDIA)
@alexandraj777 (@RubberDuckyLabs)
1
44
Big thanks to our lightning talks from
🚗 @effyyzhang (@baserunai) - Sharing the journey of going from design to LLMs & launching Friday
@ShreyaR (@guardrails_ai) - For talking about why evaluating LLMs is so hard
#llms #llmops #ml #yc #startups
1
66
Big thanks to Yao Xu for MCing the event!
Big thanks to the @FeatureformML crew (@Shabbyjooon + @simba_khadder ) for sponsoring the event & ensuring we had swag for the speakers !
#featureformml #womeninml #mlops
1
1
42
First off, big thanks to our lightning talk & panel speakers!
@effyyzhang (@baserunai)
@ShreyaR (@guardrails_ai)
@DataCereal (@VertaAI)
@biankaroy_ (@NVIDIA)
@alexandraj777 (@RubberDuckyLabs)
Yao Xu (@AWS)
#mlops #womeninml #ycombinator #yc #mlstartups
1
3
247
💡 Our blog post proposes a comprehensive framework to diagnose, strategize, and plan your ML Platform Adoption journey. Dive in! bit.ly/3OOagLq #platformengineering
15
🌐 The challenges of supporting ML in production are real, but it's never been faster or easier to deploy powerful applications. #mlengineer
1
1
84