The paradox of Data Science: You need to simultaneously believe in both miracles and cold, hard facts. That’s because DS leaders are responsible for jointly A) Seeing the game-changing future of data in their business in 18 months (getting significant data projects conception → production takes forever!) B) Being realistic about what data can do now This dichotomy can be extreme — and keeping a steady hand on the wheel is made harder by how wildly different the vibes might be amongst stakeholders at different levels. The image Hugo Bowne-Anderson shared last week in our first Data Dialogues event with Brad Klingenberg captures it beautifully in the age of AI — your C-suite might be incredibly optimistic, while folks with hands on keyboard are much less sure their project this quarter is going to work well enough to get them promoted. Data Leaders are left trying to thread the needle. Relative to many areas, optimism about data projects is surprisingly vibes-driven — I think because they take a long time and many of the approaches are just generally less proven. But it does feel ironic given we call ourselves scientists 🧐 What to do? My advice is to write down and disseminate a strategy which explicitly ladders together a portfolio of bets that span time frames. The more assertive you can be about how it all comes together, the better; the anti-pattern is to have external vibes decide it all for you. PS — Quick plug for the Data Dialogues: a private, unrecorded biweekly forum for data leaders to discuss critical and timely topics. We had a dozen senior folks last time and are expanding the group selectively. Next event is with executive coach Lilei Xu next Thursday — link to apply in the comments! #Leadership #CareerGrowth #DataScience #MachineLearning #AI
Delphina
Software Development
AI Data Scientist making classic ML step change faster + easier
About us
Delphina is an AI Data Scientist that helps data science teams drive business impact faster by automating painstaking, routine work, while also freeing partners from tedious productionization efforts. Join the waitlist at www.delphina.ai! Did we mention we're hiring? Check our job postings at https://jobs.ashbyhq.com/Delphina.
- Website
-
https://delphina.ai/
External link for Delphina
- Industry
- Software Development
- Company size
- 2-10 employees
- Headquarters
- San Francisco
- Type
- Privately Held
Locations
-
Primary
San Francisco, US
Employees at Delphina
Updates
-
As a Data Scientist: how effective are you at public speaking? Let’s be honest — most of us data geeks didn’t grow up in debate club and aren’t naturally the most eloquent in the room. It’s obvious that being trusted to present to the CEO means your leadership chain will see you as ready for bigger things. So are you currently regularly asked to speak at All Hands — or do you find yourself sidelined by your PM? Data Scientists typically have opportunities to get in front of leadership earlier and more often than other technical functions because executives want to “see the data”. Yet so many scientists don’t recognize that's a **huge** opportunity — because taking advantage of it feels unnatural and requires some work. Here are 3 things I actively do to be a crisp and dynamic speaker: 1/ Align your narrative to the arc of Problem, Solution, Benefit. It’s often hard to know where to start with a presentation, but you can nearly always use that simple arc — and it’s easy for audiences to understand and remember. “The problem was to measure the quality of recommendations; the solution is this new metric; the benefit is this new insight”. 2/ Make it real with examples. Using specific examples often feels a bit unnatural to me because I know they were cherry picked; n=1 is the definition of not stat sig. But anecdotes about real customers or events go a remarkably long way in helping the human mind understand what numbers really mean. My rule of thumb is to use more examples than feels comfortable. 3/ Practice. Serious speakers practice relentlessly — trying out talking points to themselves, to the mirror, to low stakes audiences. Even short insights need substantial polish — as Winston Churchill once said, "If you want me to speak for two minutes, it will take me three weeks of preparation. If you want me to speak for an hour, I am ready now." Becoming a strong communicator can be the fast path to significantly more influence — take advantage! Shoutout to Lauren Weinstein who has been super helpful to me in my journey as a communicator — highly recommend her classes and/or videos on YouTube. Curious what resonates, and what tips or resources have you found helpful? #datascience #machinelearning #career #leadership
-
Delphina reposted this
🚀 New Newsletter Alert! I’m thrilled to share the latest edition of my newsletter, Vanishing Gradients, where I dive into some exciting topics in the world of data science, ML, and AI. Here’s what you’ll find in this week’s edition: 1. Kicking off Data Dialogs: I’m launching a new series of conversations for data leaders, starting on August 22. Our first session features Brad Klingenberg, where we’ll explore how traditional ML is evolving with Generative AI and what that means for data science leaders. 🗓️ 2. Cutting AI Assistant Costs: We recently conducted a study where we compared CALM (Conversational AI with Language Models, Rasa) with LangChain/LangGraph. The results were compelling—CALM reduced operational costs by up to 77.8%, delivered 4x faster responses, and maintained high reliability by adhering to crucial business logic (with Alan Nichol, Daksh Varshneya). 📊 3. Podcast with Dan Becker Becker & Hamel H.: I had the chance to chat with two veterans in the AI space about their experiences teaching LLMs to thousands of data scientists. We discuss everything from the challenges of fine-tuning large language models to the future of AI education. 🎙️ 4. The AI Revolution with spaCy Creators: I did a live podcast recording about NLP and AI with Ines Montani and Matthew Honnibal from Explosion. We discussed their journey with spaCy, the role of open-source in AI, and how they’re pushing the boundaries of what’s possible in NLP. 🌐 🔗 in comment 🤗
-
-
📢 Calling all Data Science Leaders! As we're building Delphina we've been surprised to learn how many data science / ML / AI leaders are on their own when it comes to navigating hard pressing issues. They don't need what's on LinkedIn and Twitter — they need what actually works. But the reality is there just aren't many spaces for them to have open, honest dialogs. So I’m thrilled today to announce Delphina Data Dialogs — a new bi-weekly application-only online private forum where top minds in data science, ML, and AI can learn, connect, and tackle pressing challenges together. First session: > 🗓️ Thursday, August 22, 4:00 PM PT on Zoom > 🎙️ Featured speaker: Brad Klingenberg, founder of Naro and former Chief Algorithms Officer at Stitch Fix > 🎙️Moderated by Hugo Bowne-Anderson, host and creator of the Vanishing Gradients podcast > 📌 Topic: "Navigating the AI Revolution: where's the data science?" — exploring the interplay between traditional ML and GenAI (the good, the bad, and the ugly) and how it's reshaping our field. The Dialogs are space for authentic conversations. No recordings, no sales pitches — Chatham House Rules and real talk about timely topics. We’ve got a deep roster of incredible speakers to come, spanning all facets of data science strategy, tactics, and infrastructure including Lilei Xu, Min Cai, Kane Sweeney, Ali Rauh, Victor Kostyuk and more Check out the full details and apply to join us: https://lnkd.in/gsCXpz55 #DataScience #AI #MachineLearning #LeadershipForum
-
Delphina reposted this
If you ask most data leaders, “What drives your users to take the highest value actions in your product?”, they’ll gaze back at you with a pained look on their face. They’ll probably respond through gritted teeth, “That’s a hard question.” And they’re right. It’s not that they don't care; the opposite in fact. But it’s an incredibly complex puzzle, and they wish they had better answers. Throughout my career, understanding the drivers of high value actions has been *the* burning analytics question. At Wealthfront, we obsessed over what led customers to transfer their other investment accounts to us. At Uber, it was the factors behind frequent trips and subscription sign-ups. At Gopuff, it was what drove large orders and purchases of high-margin products. The problem is, traditional analytics tools like BI dashboards and spreadsheets can’t untangle the web of factors that lead to high value actions. Answering these questions requires high-dimensional causal factor analysis, decomposing outcomes across dozens, or hundreds, or even thousands of input variables. In other words, they require machine learning. This is what the most advanced analytics teams are doing — using ML to find the needles in the haystack and unveil unexpected relationships between behaviors and outcomes. The good news: upgrading your product analytics with ML is within reach. In our latest article, Jeremy and I break down three core techniques you can use today. The topic is on our mind because we’re coming across it frequently at Delphina. We're eager as always for feedback and reactions, and if you’re tackling a similar problem and want to brainstorm, reach out! #datascience #analytics #machinelearning #artificialintelligence
What advanced analytics teams are doing that you aren’t
Duncan Gilchrist on LinkedIn
-
Delphina reposted this
Love the Delphina shoutout Rob! Agents are going to change everything. And we're just getting started 🔥
There is tremendous hype around AI agents today. Hoping to separate the signal from the noise? Read on:
Agents Are The Future Of AI. Where Are The Startup Opportunities?
social-www.forbes.com
-
Delphina reposted this
I’ve interviewed or advised hundreds of new PhDs trying to break into data science in tech. Even the most accomplished academics tend to struggle in the same area: the onsite. They’re coming from a world where their papers and CV form an encyclopedic track record of their expertise to a new reality where a single 45-minute interview can decide their fate. They’ve never done this before, and they’re typically wildly unprepared. On the other side of the table, hiring managers are in a tough spot. They can’t afford to blindly disqualify talented candidates because of an awkward interview: there won’t be many candidates left! Falling prey to the cardinal sin of recruiting — hiring someone who’s good at interviewing instead of someone who’s actually good at the job — is way too easy with new PhDs. How can you cut through the noise? Over time I’ve built a short list of attributes I look for: 1/ Excitement to learn all parts of the job: have they internalized that working in tech means actually _building technology_? 2/ Evidence of multidisciplinary thinking: do they have a sledge hammer for which they must find a nail, or do they have a real toolbox? 3/ Ability to navigate multiple levels of altitude: PhDs tend to love the microscopic — but can they zoom out to see the whole ecosystem? 4/ A burning need to get things done: do they live and breathe a need for speed? 5/ Bonus: public speaking superpowers: explaining things clearly to an audience is incredibly valuable – and is a spike that’s easy to miss in standard interview loops. I unpack these in my latest post with Jeremy, including what questions to ask (and how pro job seekers respond), below. Let us know what you think, and please share: what do you look for? #hiring #leadership #datascience #machinelearning
Why PhDs whiff the onsite, and how to find a diamond in the rough
Duncan Gilchrist on LinkedIn
-
Delphina reposted this
When I was leading various data science groups at Uber there was an ever-present and uncomfortably-high likelihood that a Data Science Disaster would derail my day. It might be Thursday around 7pm; perhaps someone on a partner team noticed something looking funky in the evening rush hour of an important metro. All of a sudden, that uBeer I was about to grab would turn into a whirlwind of analysis, strained uChats, and jams to dig in. Sometimes there was a major problem, and sometimes it was nothing. At the time, I thought Uber’s culture was just intense. (That was true!) But in hindsight I’ve also realized: ML itself is fundamentally super intense. ML can be a lever that moves the world — when it works well, it delivers incredible, transformative value. But when it doesn’t work, or even when it’s just mediocre, that lever can snap and slap you in the face. ML that’s merely mediocre can be downright dangerous to the business. Unlike many business functions, the returns to quality in ML are highly nonlinear, and not-great ML is impossible to tolerate. That’s because ML’s scale is massive, it’s hard to monitor, and it’s too easy to inadvertently make mistakes. ML is also immature as a domain – many leaders don’t know problems to look for, and many data scientists don’t have the acumen to foresee where things can go wrong. Problems can burn silently for months, costing companies dearly. In ML, quality is surprisingly binary, and Jeremy and I call the area of mediocre quality the “Danger Zone”: quality that would be uninspiring but tolerable in many functions but is simply unacceptable in ML. In our latest post we unpack why, how to identify it, and how to keep your company in the clear. How have you navigated the danger zone? Let us know what you think! #datascience #machinelearning #artificialintelligence
The Danger Zone in Data Science
Duncan Gilchrist on LinkedIn
-
Delphina reposted this
Really excited for this hybrid virtual + in-person (Seattle) meetup with Yurui (Rui) Tong, Wee Hyong T. and Duncan Gilchrist. We'll be discussing how to build (or in some cases, try to build...) AI products from the ground up, and how Remitly, Delphina, Corgi Labs and Microsoft approach this, plus how the generative AI wave has helped us shorten the time from 0 to 1 (and the time for 0 to 1 recursions when the results don't turn out as you expected...) It should be an awesome discussions, and even better, it's set for a APAC, EMEA and PST friendly time to foster global inclusivity!
🚀 What does it REALLY take to deliver successful AI products? From data quality all the way to decision making! 🚀 Building AI products is like riding an emotional roller coaster - it's exciting, frustrating, and satisfying all.at.the.same.time! Come and join us for a dynamic discussion, where you'll hear great stories/insights from industry leaders and builders: Wee Hyong T. (Partner @ Microsoft | Data and AI Product Leader | Author) Saif Farooqui (Founder + CEO @ Corgi Labs - Payments AI Solutions) Duncan Gilchrist (Co-founder @ Delphina - AI Data Scientist) 🗓 Date: Wednesday, June 12th, 2024 🕕 Time: 6:00 PM to 7:30 PM PT 📍 Location: Online & Remitly’s downtown Seattle office https://lnkd.in/djqchc_K Looking forward to seeing you there and sharing all the ups and downs in building AI products! FYI, a Zoom link will be added on the Meetup page closer to the event .
What does it REALLY take to deliver successful AI products?, Wed, Jun 12, 2024, 5:30 PM | Meetup
meetup.com
-
Don't focus on building experience. Build judgment instead. How can it be that so many outrageously successful startups had founders with virtually zero experience? Think Zuck at Facebook, Gates at Microsoft, Sergey & Larry at Google. Early in my career I felt it must be mostly luck — those folks stumbled on to the right problem at the right time. Luck is certainly part of it. But as I’ve grown, I’ve realized I was thinking about it totally wrong. Judgment matters vastly more than actual experience. Judgment is the ability to read the tea leaves and make the right call. That judgment trumps experience is obviously true once you think about it — who cares if you’ve done the thing before if you can make the right decision regardless. And building judgment is obviously the only way to hyperscale your career: getting lived experiences take too much living! Founders like the ones at the top built judgment super fast. Different functions have particular kinds of judgment that are especially important. In data science, you need to build a sense of smell for when data doesn’t make sense, and also for what kinds of problems ML and AI can solve and which they can’t. If you don’t have good judgment on these issues, you won’t go far. In many ways judgment is easier to build than experience. Here are 3 things you can do: 1/ “Pretend” you have a bigger job. At Uber, I liked to play a game in my own head: Whenever I was in a review meeting with Dara, I’d ask myself, what question should he ask next? When he inevitably asked something different, I’d ask myself, why was that his question? 2/ Find a mentor who is 1-2 steps ahead of you. These folks can give you a download on their problems — you can probe the hard judgment calls they’ve made recently, how they made them, and what they learned. 3/ Read complex books. Biographies and deep novels are a great way to see lots of new, hard situations. Again, ask yourself — what would I have done in that context? Self-help / business books are less useful, since they abstract away complexity in favor of simple takeaways. Would love to hear what resonates, and what other tricks I’m missing? #datascience #machinelearning #jobs