StarZero All shows Try the podcast clipper →

Jan 16, 2014 · 23m · mad

Tim Hwang, FiscalNote // Data Driven NYC 21 // Dec 2013 (Hosted by FirstMark Capital)

Tim Hwang · 12m spoken Matt Turck · 50s spoken Rob Rakowski · 29s spoken Larry Schiller · 20s spoken Mike Rosen · 18s spoken Jonathan Dayton · 15s spoken
0:00 / 0:00
▶ Watch on YouTube →

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

At Data Driven NYC, Tim Hwang and Vlad from FiscalNote present their data science platform, which utilizes machine learning and natural language processing to transform unstructured government data into predictive insights for policy analysis. They detail the technical pipeline, market applications across financial and enterprise sectors, and backtesting performance during an interactive Q&A.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Matt holds 4.4% of the talking time here. How this is scored →

Matt as informed peer 0.6 Guest teaching 1.6 Guest disagreement 0.2 Matt pushing back 0.6
05100:0010:0020:000:42–3:32 · Matt as informed peer 0/10 The Unstructured Government Data Dilemma This is a monologue presentation by guest Tim Hwang outlining unstructured government data challenges. The host does not speak or engage during this presentation segment.3:32–9:13 · Matt as informed peer 0/10 Introducing FiscalNote Prophecy and Predictive Analytics Technical presentation segment delivered by guest Vlad detailing machine learning classifiers and data acquisition. The host remains silent throughout this presentation block.9:13–15:28 · Matt as informed peer 0/10 Categorizing Legislation Across Industries and Subcategories Tim Hwang completes the main presentation discussing NLP nuances and business relationships. Host Matt Turck has no dialogue during the pitch.15:28–18:32 · Matt as informed peer 1/10 Q&A: Target Customers and Legal Application Separation Matt Turck moderates Q&A, introducing audience questions about target customer categories. The interaction is collaborative and informational with zero host pushback or guest combativeness.18:32–23:22 · Matt as informed peer 2/10 Q&A: Backtesting, External Datasets, and Model Variables Matt Turck asks an insightful question about state vs. federal market dynamics and predictability. Guest Tim Hwang reframes baseline accuracy concerns when answering audience questions politely.0:42–3:32 · Guest teaching 1/10 The Unstructured Government Data Dilemma This is a monologue presentation by guest Tim Hwang outlining unstructured government data challenges. The host does not speak or engage during this presentation segment.3:32–9:13 · Guest teaching 1/10 Introducing FiscalNote Prophecy and Predictive Analytics Technical presentation segment delivered by guest Vlad detailing machine learning classifiers and data acquisition. The host remains silent throughout this presentation block.9:13–15:28 · Guest teaching 1/10 Categorizing Legislation Across Industries and Subcategories Tim Hwang completes the main presentation discussing NLP nuances and business relationships. Host Matt Turck has no dialogue during the pitch.15:28–18:32 · Guest teaching 2/10 Q&A: Target Customers and Legal Application Separation Matt Turck moderates Q&A, introducing audience questions about target customer categories. The interaction is collaborative and informational with zero host pushback or guest combativeness.18:32–23:22 · Guest teaching 3/10 Q&A: Backtesting, External Datasets, and Model Variables Matt Turck asks an insightful question about state vs. federal market dynamics and predictability. Guest Tim Hwang reframes baseline accuracy concerns when answering audience questions politely.0:42–3:32 · Guest disagreement 0/10 The Unstructured Government Data Dilemma This is a monologue presentation by guest Tim Hwang outlining unstructured government data challenges. The host does not speak or engage during this presentation segment.3:32–9:13 · Guest disagreement 0/10 Introducing FiscalNote Prophecy and Predictive Analytics Technical presentation segment delivered by guest Vlad detailing machine learning classifiers and data acquisition. The host remains silent throughout this presentation block.9:13–15:28 · Guest disagreement 0/10 Categorizing Legislation Across Industries and Subcategories Tim Hwang completes the main presentation discussing NLP nuances and business relationships. Host Matt Turck has no dialogue during the pitch.15:28–18:32 · Guest disagreement 0/10 Q&A: Target Customers and Legal Application Separation Matt Turck moderates Q&A, introducing audience questions about target customer categories. The interaction is collaborative and informational with zero host pushback or guest combativeness.18:32–23:22 · Guest disagreement 1/10 Q&A: Backtesting, External Datasets, and Model Variables Matt Turck asks an insightful question about state vs. federal market dynamics and predictability. Guest Tim Hwang reframes baseline accuracy concerns when answering audience questions politely.0:42–3:32 · Matt pushing back 0/10 The Unstructured Government Data Dilemma This is a monologue presentation by guest Tim Hwang outlining unstructured government data challenges. The host does not speak or engage during this presentation segment.3:32–9:13 · Matt pushing back 0/10 Introducing FiscalNote Prophecy and Predictive Analytics Technical presentation segment delivered by guest Vlad detailing machine learning classifiers and data acquisition. The host remains silent throughout this presentation block.9:13–15:28 · Matt pushing back 0/10 Categorizing Legislation Across Industries and Subcategories Tim Hwang completes the main presentation discussing NLP nuances and business relationships. Host Matt Turck has no dialogue during the pitch.15:28–18:32 · Matt pushing back 1/10 Q&A: Target Customers and Legal Application Separation Matt Turck moderates Q&A, introducing audience questions about target customer categories. The interaction is collaborative and informational with zero host pushback or guest combativeness.18:32–23:22 · Matt pushing back 2/10 Q&A: Backtesting, External Datasets, and Model Variables Matt Turck asks an insightful question about state vs. federal market dynamics and predictability. Guest Tim Hwang reframes baseline accuracy concerns when answering audience questions politely.

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

0:00 · Matt 16.9% · guest 83.1%0:00 · Matt 16.9% · guest 83.1%3:00 · Matt 0% · guest 100%3:00 · Matt 0% · guest 100%6:00 · Matt 0% · guest 100%6:00 · Matt 0% · guest 100%9:00 · Matt 0% · guest 100%9:00 · Matt 0% · guest 100%12:00 · Matt 0% · guest 100%12:00 · Matt 0% · guest 100%15:00 · Matt 7% · guest 93%15:00 · Matt 7% · guest 93%18:00 · Matt 7.4% · guest 92.6%18:00 · Matt 7.4% · guest 92.6%21:00 · Matt 4.7% · guest 95.3%21:00 · Matt 4.7% · guest 95.3%
Sharpest disagreement ▶ 21:50 Reframing baseline passage accuracy

Tim gently counters the audience member's premise about high baseline failures by explaining how prediction algorithms surface early-stage bills across varying state dynamics.

 Hardest push from Matt ▶ 20:43 Challenging state focus rationale

Host Matt Turck challenges the guest on whether their focus on state data is driven by lower market competition or higher inherent structural predictability.

 Biggest teaching moment ▶ 13:00 Syntax nuance in legal text analysis

Tim Hwang educates the audience on how subtle punctuation shifts in legal text drastically reframe regulatory meaning and algorithmic prediction models.

 Matt holds his own ▶ 20:43 Framing market dynamics

Host Matt Turck displays domain expertise by positing a sharp conceptual distinction between competitive landscape gaps and statistical predictability in state legislation.
the scores for every segment, with the reasoning behind each
ChapterTopicMatt as informed peerGuest teachingGuest disagreementMatt pushing backWhy
The Unstructured Government Data Dilemma 0100 This is a monologue presentation by guest Tim Hwang outlining unstructured government data challenges. The host does not speak or engage during this presentation segment.
Introducing FiscalNote Prophecy and Predictive Analytics 0100 Technical presentation segment delivered by guest Vlad detailing machine learning classifiers and data acquisition. The host remains silent throughout this presentation block.
Categorizing Legislation Across Industries and Subcategories 0100 Tim Hwang completes the main presentation discussing NLP nuances and business relationships. Host Matt Turck has no dialogue during the pitch.
Q&A: Target Customers and Legal Application Separation 1201 Matt Turck moderates Q&A, introducing audience questions about target customer categories. The interaction is collaborative and informational with zero host pushback or guest combativeness.
Q&A: Backtesting, External Datasets, and Model Variables 2312 Matt Turck asks an insightful question about state vs. federal market dynamics and predictability. Guest Tim Hwang reframes baseline accuracy concerns when answering audience questions politely.

Statements from this episode (5)

Assertion Not checkable as stated
Tim Hwang: 90% to 95% of government data is unstructured
“90 to 95% of government data is unstructured.”
Tim Hwang Jan 16, 2014 ▶ 1:09
Assertion Not checkable as stated
FiscalNote's prediction algorithms determine bill passage with 94% accuracy
“Currently, our prediction algorithms are about 94% accurate in being able to determine at first reader whether or not a bill is going to pass.”
Tim Hwang Jan 16, 2014 ▶ 4:22
Assertion Not checkable as stated
Tim Hwang states state legislation impacts $4 trillion in US spending
“So, state legislation obviously impacts almost four trillion dollars in in, ah, in our economy spending today.”
Tim Hwang Jan 16, 2014 ▶ 5:59
Prediction Not checkable as stated
Tim Hwang predicts AI models will enable automated 'high-frequency litigation'
“In the future we might have something like, ah, high frequency litigation, and that's, I don't think that's too far from the truth in terms of being able to understand and, ah, predict these types of outcomes based off of, ah, past datasets.”
Tim Hwang Jan 16, 2014 ▶ 17:51
Assertion Partly supported
Tim Hwang says roughly 94% of New York state legislative bills fail
“In New York, I think something like 93 or 94% of bills end up failing you know, by the time they get to committee.”
Tim Hwang Jan 16, 2014 ▶ 22:06
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 400 conversations 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