StarZero All shows Try the podcast clipper →

Jul 16, 2017 · 20m · top-founders

722: This Machine Learning Agency did $800k Last Year

Michael Segala · 11m spoken Nathan Latka · 7m spoken
0:00 / 0:00

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

In this episode of The Top, Nathan Latka interviews Michael Segala, co-founder and CEO of SFL Scientific, exploring how three former CERN particle physicists bootstrapped a bespoke AI and machine learning consulting firm from $2,000 to over $800,000 in annual revenue.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Nathan holds 41.2% of the talking time here. How this is scored →

Nathan as informed peer 3.6 Guest teaching 3.3 Guest disagreement 1.4 Nathan pushing back 3.4
05100:0010:0020:001:07–5:15 · Nathan as informed peer 3/10 Michael Segala's Background in Particle Physics and Data Science Nathan begins by challenging Michael on whether SFL Scientific is the 'real deal' versus hype. Michael educates him on the team's particle physics background and CERN/LHC research, leaving Nathan to admit he is not that scientific.5:15–9:15 · Nathan as informed peer 4/10 Consulting Pricing Strategies, Scoping, and Founder Equity Distribution Nathan drills down on SFL's consulting pricing mechanics and equity split with role-playing questions. Michael explains their time-and-materials strategy, scoping process, and how they provide clear business ROI.9:15–11:56 · Nathan as informed peer 6/10 Reinvestment Strategy, Revenue Scaling, and Managing Concentration Risk Nathan presses Michael on founder pay, revenue growth, and client concentration risk (>20% in one client). Michael acknowledges the risk transparently and details their mitigation tactics.11:57–14:05 · Nathan as informed peer 5/10 Cross-Industry Diversification and the Universality of Data Science Nathan challenges how Michael can win against niche vertical agencies. Michael counters by schooling Nathan on how mathematical problems in data science are structurally identical across pharma, insurance, and tech.14:05–17:49 · Nathan as informed peer 6/10 Knowledge Graphs, Data Integrity, and Breakthroughs in Healthcare AI Nathan cites systems thinking and network effect loops, probing how SFL creates business moats and handles client exclusivity. Michael validates the concept through knowledge graphs and discusses high-impact applications in healthcare.17:49–20:04 · Nathan as informed peer 0/10 Nathan Latka's Exclusive SaaS Database Announcement at GetLatka.com Host solo mid-roll ad pitch promoting the GetLatka SaaS database. No guest interaction.20:04–20:48 · Nathan as informed peer 1/10 Episode Conclusion and SFL Scientific Performance Recap Standard Famous Five lightning round and episode wrap-up summary by the host.1:07–5:15 · Guest teaching 6/10 Michael Segala's Background in Particle Physics and Data Science Nathan begins by challenging Michael on whether SFL Scientific is the 'real deal' versus hype. Michael educates him on the team's particle physics background and CERN/LHC research, leaving Nathan to admit he is not that scientific.5:15–9:15 · Guest teaching 3/10 Consulting Pricing Strategies, Scoping, and Founder Equity Distribution Nathan drills down on SFL's consulting pricing mechanics and equity split with role-playing questions. Michael explains their time-and-materials strategy, scoping process, and how they provide clear business ROI.9:15–11:56 · Guest teaching 3/10 Reinvestment Strategy, Revenue Scaling, and Managing Concentration Risk Nathan presses Michael on founder pay, revenue growth, and client concentration risk (>20% in one client). Michael acknowledges the risk transparently and details their mitigation tactics.11:57–14:05 · Guest teaching 6/10 Cross-Industry Diversification and the Universality of Data Science Nathan challenges how Michael can win against niche vertical agencies. Michael counters by schooling Nathan on how mathematical problems in data science are structurally identical across pharma, insurance, and tech.14:05–17:49 · Guest teaching 4/10 Knowledge Graphs, Data Integrity, and Breakthroughs in Healthcare AI Nathan cites systems thinking and network effect loops, probing how SFL creates business moats and handles client exclusivity. Michael validates the concept through knowledge graphs and discusses high-impact applications in healthcare.17:49–20:04 · Guest teaching 0/10 Nathan Latka's Exclusive SaaS Database Announcement at GetLatka.com Host solo mid-roll ad pitch promoting the GetLatka SaaS database. No guest interaction.20:04–20:48 · Guest teaching 1/10 Episode Conclusion and SFL Scientific Performance Recap Standard Famous Five lightning round and episode wrap-up summary by the host.1:07–5:15 · Guest disagreement 2/10 Michael Segala's Background in Particle Physics and Data Science Nathan begins by challenging Michael on whether SFL Scientific is the 'real deal' versus hype. Michael educates him on the team's particle physics background and CERN/LHC research, leaving Nathan to admit he is not that scientific.5:15–9:15 · Guest disagreement 1/10 Consulting Pricing Strategies, Scoping, and Founder Equity Distribution Nathan drills down on SFL's consulting pricing mechanics and equity split with role-playing questions. Michael explains their time-and-materials strategy, scoping process, and how they provide clear business ROI.9:15–11:56 · Guest disagreement 2/10 Reinvestment Strategy, Revenue Scaling, and Managing Concentration Risk Nathan presses Michael on founder pay, revenue growth, and client concentration risk (>20% in one client). Michael acknowledges the risk transparently and details their mitigation tactics.11:57–14:05 · Guest disagreement 3/10 Cross-Industry Diversification and the Universality of Data Science Nathan challenges how Michael can win against niche vertical agencies. Michael counters by schooling Nathan on how mathematical problems in data science are structurally identical across pharma, insurance, and tech.14:05–17:49 · Guest disagreement 2/10 Knowledge Graphs, Data Integrity, and Breakthroughs in Healthcare AI Nathan cites systems thinking and network effect loops, probing how SFL creates business moats and handles client exclusivity. Michael validates the concept through knowledge graphs and discusses high-impact applications in healthcare.17:49–20:04 · Guest disagreement 0/10 Nathan Latka's Exclusive SaaS Database Announcement at GetLatka.com Host solo mid-roll ad pitch promoting the GetLatka SaaS database. No guest interaction.20:04–20:48 · Guest disagreement 0/10 Episode Conclusion and SFL Scientific Performance Recap Standard Famous Five lightning round and episode wrap-up summary by the host.1:07–5:15 · Nathan pushing back 4/10 Michael Segala's Background in Particle Physics and Data Science Nathan begins by challenging Michael on whether SFL Scientific is the 'real deal' versus hype. Michael educates him on the team's particle physics background and CERN/LHC research, leaving Nathan to admit he is not that scientific.5:15–9:15 · Nathan pushing back 4/10 Consulting Pricing Strategies, Scoping, and Founder Equity Distribution Nathan drills down on SFL's consulting pricing mechanics and equity split with role-playing questions. Michael explains their time-and-materials strategy, scoping process, and how they provide clear business ROI.9:15–11:56 · Nathan pushing back 6/10 Reinvestment Strategy, Revenue Scaling, and Managing Concentration Risk Nathan presses Michael on founder pay, revenue growth, and client concentration risk (>20% in one client). Michael acknowledges the risk transparently and details their mitigation tactics.11:57–14:05 · Nathan pushing back 5/10 Cross-Industry Diversification and the Universality of Data Science Nathan challenges how Michael can win against niche vertical agencies. Michael counters by schooling Nathan on how mathematical problems in data science are structurally identical across pharma, insurance, and tech.14:05–17:49 · Nathan pushing back 5/10 Knowledge Graphs, Data Integrity, and Breakthroughs in Healthcare AI Nathan cites systems thinking and network effect loops, probing how SFL creates business moats and handles client exclusivity. Michael validates the concept through knowledge graphs and discusses high-impact applications in healthcare.17:49–20:04 · Nathan pushing back 0/10 Nathan Latka's Exclusive SaaS Database Announcement at GetLatka.com Host solo mid-roll ad pitch promoting the GetLatka SaaS database. No guest interaction.20:04–20:48 · Nathan pushing back 0/10 Episode Conclusion and SFL Scientific Performance Recap Standard Famous Five lightning round and episode wrap-up summary by the host.

speaking balance: gold is Nathan, purple is the guest (3 minute bins)

0:00 · Nathan 68% · guest 32%0:00 · Nathan 68% · guest 32%3:00 · Nathan 18.1% · guest 81.9%3:00 · Nathan 18.1% · guest 81.9%6:00 · Nathan 25.4% · guest 74.6%6:00 · Nathan 25.4% · guest 74.6%9:00 · Nathan 42.6% · guest 57.4%9:00 · Nathan 42.6% · guest 57.4%12:00 · Nathan 24.4% · guest 75.6%12:00 · Nathan 24.4% · guest 75.6%15:00 · Nathan 27.2% · guest 72.8%15:00 · Nathan 27.2% · guest 72.8%18:00 · Nathan 86.7% · guest 13.3%18:00 · Nathan 86.7% · guest 13.3%
Sharpest disagreement ▶ 13:04 Michael rejects vertical agency advantage

Michael dismisses the notion that domain-specialized agencies have an edge, asserting that data science challenges across disparate industries are practically identical.

 Hardest push from Nathan ▶ 11:38 Nathan presses on revenue concentration risk

Nathan bluntly corners Michael on whether a single client accounts for over 20% of revenue and warns of the dangers of agency layoffs.

 Biggest teaching moment ▶ 1:53 Michael clarifies CERN Large Hadron Collider experience

When Nathan asks if LHC is an exam, Michael clarifies that it is CERN's particle collider and explains how subatomic physics R&D translates into rigorous commercial data science.

 Nathan holds their own ▶ 14:05 Nathan introduces systems thinking and reinforcing feedback loops

Nathan references 'Thinking in Systems' and challenges Michael to explain how machine learning builds defensible data moats and network effects.
the scores for every segment, with the reasoning behind each
ChapterTopicNathan as informed peerGuest teachingGuest disagreementNathan pushing backWhy
Michael Segala's Background in Particle Physics and Data Science 3624 Nathan begins by challenging Michael on whether SFL Scientific is the 'real deal' versus hype. Michael educates him on the team's particle physics background and CERN/LHC research, leaving Nathan to admit he is not that scientific.
Consulting Pricing Strategies, Scoping, and Founder Equity Distribution 4314 Nathan drills down on SFL's consulting pricing mechanics and equity split with role-playing questions. Michael explains their time-and-materials strategy, scoping process, and how they provide clear business ROI.
Reinvestment Strategy, Revenue Scaling, and Managing Concentration Risk 6326 Nathan presses Michael on founder pay, revenue growth, and client concentration risk (>20% in one client). Michael acknowledges the risk transparently and details their mitigation tactics.
Cross-Industry Diversification and the Universality of Data Science 5635 Nathan challenges how Michael can win against niche vertical agencies. Michael counters by schooling Nathan on how mathematical problems in data science are structurally identical across pharma, insurance, and tech.
Knowledge Graphs, Data Integrity, and Breakthroughs in Healthcare AI 6425 Nathan cites systems thinking and network effect loops, probing how SFL creates business moats and handles client exclusivity. Michael validates the concept through knowledge graphs and discusses high-impact applications in healthcare.
Nathan Latka's Exclusive SaaS Database Announcement at GetLatka.com 0000 Host solo mid-roll ad pitch promoting the GetLatka SaaS database. No guest interaction.
Episode Conclusion and SFL Scientific Performance Recap 1100 Standard Famous Five lightning round and episode wrap-up summary by the host.

Statements from this episode (14)

Assertion Supported
Segala: SFL Scientific founding team are all particle physics PhDs
“So our background, at least from the founding team, we were all particle physicists doing our PhDs in very kind of cutting edge R&D space.”
Michael Segala Jul 16, 2017 ▶ 1:59
Disclosure
Segala: SFL Scientific is completely bootstrapped
“We're completely bootstrapped.”
Michael Segala Jul 16, 2017 ▶ 2:56
Assertion Not checkable as stated
Segala: SFL Scientific started with about $2,000
“I think all in, we were about 2000 dollars, which is pretty incredible.”
Michael Segala Jul 16, 2017 ▶ 3:01
Assertion Contradicted
Segala: SFL Scientific's ML suite got Stanford client FDA approval
“So they basically hired us to build out this entire suite of AI machine learning product solutions, and we did that, and it actually got them through FDA regulations, right?”
Michael Segala Jul 16, 2017 ▶ 3:58
Disclosure
SFL Scientific's starting client contracts range from $50,000 to over $100,000
“High five figure, low six figures is usually the typical kind of starting point for clients.”
Michael Segala Jul 16, 2017 ▶ 6:03
Insight
Segala: Most clients lack understanding to scope AI projects accurately
“To be honest, most clients, Aren't informed enough to understand the true scope of the project.”
Michael Segala Jul 16, 2017 ▶ 7:09
Disclosure
Segala: SFL Scientific generated low six figures in first-year revenue
“It was in the low six figures, which was enough to support the three founders.”
Michael Segala Jul 16, 2017 ▶ 8:10
Disclosure
Segala: SFL Scientific co-founders split equity 34-33-33
“Well, I took an extra percentage point, but we split it. 33, 33, 34.”
Michael Segala Jul 16, 2017 ▶ 9:10
Assertion Not checkable as stated
Segala: SFL Scientific quadrupled its revenue in 2016
“We quadrupled from 2015, we quadrupled revenue, and this year we hope to quadruple revenue again.”
Michael Segala Jul 16, 2017 ▶ 10:28
Disclosure
Segala: One client accounts for over 20% of SFL Scientific's revenue
“There is, and that's dangerous.”
Michael Segala Jul 16, 2017 ▶ 11:39
Assertion Not checkable as stated
Segala: SFL Scientific has never had a customer churn or cut contracts
“So the beautiful thing is, not to say it will never happen, so we've never had a customer leave us yet, or cut us.”
Michael Segala Jul 16, 2017 ▶ 11:57
Insight
Segala: Data science challenges across different industry verticals are fundamentally identical
“So when you look at data science in general, Across any vertical, doesn't matter, you can pick a vertical, the problems or the challenges, they don't like when you say problems, the challenges that we're solving are unanimous, right? If you're looking at healt…”
Michael Segala Jul 16, 2017 ▶ 13:11
Prediction Not checkable as stated
Segala: Knowledge graphs will revolutionize all fields in next few years
“Building these large graphs of information and knowledge that will completely revolutionize all fields in a very, you know, in the next few years.”
Michael Segala Jul 16, 2017 ▶ 15:16
Disclosure
SFL Scientific is building AI for cancer detection and 3D organ printing
“Health, for sure. We're doing some phenomenal work. I mean, from things like detecting cancer in images to actually taking reconstructed organs that we are able to detect in our algorithms, grow those organs in a kind of three-D printing environment, and then …”
Michael Segala Jul 16, 2017 ▶ 17:22
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 2,600 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more