Feb 21, 2016 · 24m · mad

A Recipe-Based Approach to Parsing // Hicham Oudghiri, Enigma (Hosted by FirstMark Capital)

Hicham Oudghiri · 20m spoken Matt Turck · 19s 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 a DataDrivenNYC event hosted by FirstMark Capital, Enigma Founder and CEO Hicham Oudghiri presents Enigma's approach to unifying fragmented public and enterprise datasets. He explains how Enigma transitioned from raw data exploration to contextual signal generation, while introducing ParseKit, a standardized DSL that streamlines complex data ingestion workflows.

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 1.6% of the talking time here. How this is scored →

Matt as informed peer 0.6 Guest teaching 4.0 Guest disagreement 0.6 Matt pushing back 0.3
05100:0010:0020:001:34–4:59 · Matt as informed peer 0/10 Public Data Explorer and Search Capabilities Hicham presents Enigma's core product features and search capabilities without any host involvement. He explains the shift from searching metadata to searching actual data points within tables. Host scores are zero as Matt Turck does not participate in this monologue.4:59–8:08 · Matt as informed peer 0/10 Data Silos and the Importance of Linking Hicham outlines the problem with siloed enterprise data systems and why linking datasets is the central challenge. He mildly critiques the broader analytics industry for missing this fundamental problem. The host does not interject or ask questions during this presentation segment.8:08–10:59 · Matt as informed peer 0/10 ETL Challenges and the Creation of ParseKit Hicham addresses ETL scale challenges and details how Enigma replaced their legacy Perl scripts with ParseKit, a DSL inspired by SysOps tooling like Chef and Ansible. The presentation remains entirely one-sided with no host involvement.10:59–13:59 · Matt as informed peer 0/10 ParseKit Live Code Walkthrough and Breakpoint Demo Hicham conducts a live walkthrough of ParseKit syntax and demonstrates a live breakpoint feature during data parsing. He lightheartedly shares an anecdote about stress during their TechCrunch Disrupt presentation. The host does not speak.13:59–16:05 · Matt as informed peer 0/10 Enterprise Platform UI and Parser Metrics Hicham concludes his presentation on the enterprise platform interface and parser metrics, emphasizing that operational code will increasingly resemble data. The host remains silent throughout the entire segment.16:05–20:08 · Matt as informed peer 3/10 Q&A: Company Growth and Open Source Plans Matt Turck opens Q&A by asking about company milestones since 2013, cracking a quick joke about VC funding when Hicham mentions his wife is a rug dealer. Matt also interjects during an audience question to playfully probe for an open-source announcement.20:08–24:50 · Matt as informed peer 1/10 Q&A: Data Skepticism and Schema Normalization Audience members ask detailed technical questions regarding data skepticism and normalization across local jurisdictions. Hicham explains how Enigma handles schema variations like Michigan's hundreds of liquor license types, while host Matt Turck only intervenes to thank the speaker and conclude.1:34–4:59 · Guest teaching 4/10 Public Data Explorer and Search Capabilities Hicham presents Enigma's core product features and search capabilities without any host involvement. He explains the shift from searching metadata to searching actual data points within tables. Host scores are zero as Matt Turck does not participate in this monologue.4:59–8:08 · Guest teaching 5/10 Data Silos and the Importance of Linking Hicham outlines the problem with siloed enterprise data systems and why linking datasets is the central challenge. He mildly critiques the broader analytics industry for missing this fundamental problem. The host does not interject or ask questions during this presentation segment.8:08–10:59 · Guest teaching 5/10 ETL Challenges and the Creation of ParseKit Hicham addresses ETL scale challenges and details how Enigma replaced their legacy Perl scripts with ParseKit, a DSL inspired by SysOps tooling like Chef and Ansible. The presentation remains entirely one-sided with no host involvement.10:59–13:59 · Guest teaching 4/10 ParseKit Live Code Walkthrough and Breakpoint Demo Hicham conducts a live walkthrough of ParseKit syntax and demonstrates a live breakpoint feature during data parsing. He lightheartedly shares an anecdote about stress during their TechCrunch Disrupt presentation. The host does not speak.13:59–16:05 · Guest teaching 3/10 Enterprise Platform UI and Parser Metrics Hicham concludes his presentation on the enterprise platform interface and parser metrics, emphasizing that operational code will increasingly resemble data. The host remains silent throughout the entire segment.16:05–20:08 · Guest teaching 2/10 Q&A: Company Growth and Open Source Plans Matt Turck opens Q&A by asking about company milestones since 2013, cracking a quick joke about VC funding when Hicham mentions his wife is a rug dealer. Matt also interjects during an audience question to playfully probe for an open-source announcement.20:08–24:50 · Guest teaching 5/10 Q&A: Data Skepticism and Schema Normalization Audience members ask detailed technical questions regarding data skepticism and normalization across local jurisdictions. Hicham explains how Enigma handles schema variations like Michigan's hundreds of liquor license types, while host Matt Turck only intervenes to thank the speaker and conclude.1:34–4:59 · Guest disagreement 0/10 Public Data Explorer and Search Capabilities Hicham presents Enigma's core product features and search capabilities without any host involvement. He explains the shift from searching metadata to searching actual data points within tables. Host scores are zero as Matt Turck does not participate in this monologue.4:59–8:08 · Guest disagreement 1/10 Data Silos and the Importance of Linking Hicham outlines the problem with siloed enterprise data systems and why linking datasets is the central challenge. He mildly critiques the broader analytics industry for missing this fundamental problem. The host does not interject or ask questions during this presentation segment.8:08–10:59 · Guest disagreement 0/10 ETL Challenges and the Creation of ParseKit Hicham addresses ETL scale challenges and details how Enigma replaced their legacy Perl scripts with ParseKit, a DSL inspired by SysOps tooling like Chef and Ansible. The presentation remains entirely one-sided with no host involvement.10:59–13:59 · Guest disagreement 0/10 ParseKit Live Code Walkthrough and Breakpoint Demo Hicham conducts a live walkthrough of ParseKit syntax and demonstrates a live breakpoint feature during data parsing. He lightheartedly shares an anecdote about stress during their TechCrunch Disrupt presentation. The host does not speak.13:59–16:05 · Guest disagreement 0/10 Enterprise Platform UI and Parser Metrics Hicham concludes his presentation on the enterprise platform interface and parser metrics, emphasizing that operational code will increasingly resemble data. The host remains silent throughout the entire segment.16:05–20:08 · Guest disagreement 1/10 Q&A: Company Growth and Open Source Plans Matt Turck opens Q&A by asking about company milestones since 2013, cracking a quick joke about VC funding when Hicham mentions his wife is a rug dealer. Matt also interjects during an audience question to playfully probe for an open-source announcement.20:08–24:50 · Guest disagreement 2/10 Q&A: Data Skepticism and Schema Normalization Audience members ask detailed technical questions regarding data skepticism and normalization across local jurisdictions. Hicham explains how Enigma handles schema variations like Michigan's hundreds of liquor license types, while host Matt Turck only intervenes to thank the speaker and conclude.1:34–4:59 · Matt pushing back 0/10 Public Data Explorer and Search Capabilities Hicham presents Enigma's core product features and search capabilities without any host involvement. He explains the shift from searching metadata to searching actual data points within tables. Host scores are zero as Matt Turck does not participate in this monologue.4:59–8:08 · Matt pushing back 0/10 Data Silos and the Importance of Linking Hicham outlines the problem with siloed enterprise data systems and why linking datasets is the central challenge. He mildly critiques the broader analytics industry for missing this fundamental problem. The host does not interject or ask questions during this presentation segment.8:08–10:59 · Matt pushing back 0/10 ETL Challenges and the Creation of ParseKit Hicham addresses ETL scale challenges and details how Enigma replaced their legacy Perl scripts with ParseKit, a DSL inspired by SysOps tooling like Chef and Ansible. The presentation remains entirely one-sided with no host involvement.10:59–13:59 · Matt pushing back 0/10 ParseKit Live Code Walkthrough and Breakpoint Demo Hicham conducts a live walkthrough of ParseKit syntax and demonstrates a live breakpoint feature during data parsing. He lightheartedly shares an anecdote about stress during their TechCrunch Disrupt presentation. The host does not speak.13:59–16:05 · Matt pushing back 0/10 Enterprise Platform UI and Parser Metrics Hicham concludes his presentation on the enterprise platform interface and parser metrics, emphasizing that operational code will increasingly resemble data. The host remains silent throughout the entire segment.16:05–20:08 · Matt pushing back 2/10 Q&A: Company Growth and Open Source Plans Matt Turck opens Q&A by asking about company milestones since 2013, cracking a quick joke about VC funding when Hicham mentions his wife is a rug dealer. Matt also interjects during an audience question to playfully probe for an open-source announcement.20:08–24:50 · Matt pushing back 0/10 Q&A: Data Skepticism and Schema Normalization Audience members ask detailed technical questions regarding data skepticism and normalization across local jurisdictions. Hicham explains how Enigma handles schema variations like Michigan's hundreds of liquor license types, while host Matt Turck only intervenes to thank the speaker and conclude.

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

0:00 · Matt 0% · guest 100%0:00 · Matt 0% · guest 100%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 10.5% · guest 89.5%15:00 · Matt 10.5% · guest 89.5%18:00 · Matt 2.3% · guest 97.7%18:00 · Matt 2.3% · guest 97.7%21:00 · Matt 0% · guest 100%21:00 · Matt 0% · guest 100%24:00 · Matt 3.3% · guest 96.7%24:00 · Matt 3.3% · guest 96.7%
Sharpest disagreement ▶ 21:05 Rejecting model infallibility

Hicham firmly counters a common industry attitude, arguing that data companies must maintain skepticism and accept that their own models are wrong rather than assuming reality is wrong.

Hardest push from Matt ▶ 19:01 Pressing for open-source announcement

Matt Turck jumps in during an audience member's question about an empty PyPy repository to directly put Hicham on the spot about making an open-source announcement.

Biggest teaching moment ▶ 22:35 Michigan liquor licenses normalization lesson

Hicham illustrates the friction of public data normalization by contrasting states with three simple liquor license categories against Michigan's hyper-specific, time-restricted hundreds of license classes.

Matt holds his own ▶ 18:05 Joking about VC funding vs rug business

Matt Turck quickly connects Hicham's comment about his wife's lucrative Moroccan rug dealership into a witty banter point about avoiding the need for venture capital.

the scores for every segment, with the reasoning behind each
ChapterTopicMatt as informed peerGuest teachingGuest disagreementMatt pushing backWhy
Public Data Explorer and Search Capabilities 0400 Hicham presents Enigma's core product features and search capabilities without any host involvement. He explains the shift from searching metadata to searching actual data points within tables. Host scores are zero as Matt Turck does not participate in this monologue.
Data Silos and the Importance of Linking 0510 Hicham outlines the problem with siloed enterprise data systems and why linking datasets is the central challenge. He mildly critiques the broader analytics industry for missing this fundamental problem. The host does not interject or ask questions during this presentation segment.
ETL Challenges and the Creation of ParseKit 0500 Hicham addresses ETL scale challenges and details how Enigma replaced their legacy Perl scripts with ParseKit, a DSL inspired by SysOps tooling like Chef and Ansible. The presentation remains entirely one-sided with no host involvement.
ParseKit Live Code Walkthrough and Breakpoint Demo 0400 Hicham conducts a live walkthrough of ParseKit syntax and demonstrates a live breakpoint feature during data parsing. He lightheartedly shares an anecdote about stress during their TechCrunch Disrupt presentation. The host does not speak.
Enterprise Platform UI and Parser Metrics 0300 Hicham concludes his presentation on the enterprise platform interface and parser metrics, emphasizing that operational code will increasingly resemble data. The host remains silent throughout the entire segment.
Q&A: Company Growth and Open Source Plans 3212 Matt Turck opens Q&A by asking about company milestones since 2013, cracking a quick joke about VC funding when Hicham mentions his wife is a rug dealer. Matt also interjects during an audience question to playfully probe for an open-source announcement.
Q&A: Data Skepticism and Schema Normalization 1520 Audience members ask detailed technical questions regarding data skepticism and normalization across local jurisdictions. Hicham explains how Enigma handles schema variations like Michigan's hundreds of liquor license types, while host Matt Turck only intervenes to thank the speaker and conclude.

Statements from this episode (7)

Assertion Not checkable as stated
Oudghiri: Enigma built first data catalog searching table rows, not metadata
“It was the first data catalog that came out with this notion of actually searching through billions of records, searching through big data, so to speak, and actually hitting on actual data points inside the tables as opposed to the metadata in the table, right…”
Hicham Oudghiri Feb 21, 2016 ▶ 1:52
Disclosure
Enigma deployed its Smoke Signals risk assessment tool with New Orleans
“And this is something that we deployed You know, ah, with the city of New Orleans, ah, have people actually go out and, ah, kind of knock on doors and deploy smoke detectives this way.”
Hicham Oudghiri Feb 21, 2016 ▶ 4:35
Insight
Oudghiri: Enterprise data problems stem from disparate contexts, not analysis limits
“Most of the problems exist not because we don't know how to analyze the data or we don't necessarily see the patterns of the data. It's just because these data sets are born in different contexts and have different purposes, right?”
Hicham Oudghiri Feb 21, 2016 ▶ 5:10
Disclosure
Enigma has over 100,000 datasets in its system
“It was just not hitting the scale that we needed to acquire, you know, the 100,000 Plus data sets that we have in the system, and kind of maintain them, and it was just not operational enough.”
Hicham Oudghiri Feb 21, 2016 ▶ 8:41
Assertion Contradicted
Oudghiri: Enigma was the only live demo at TechCrunch Disrupt
“We were the only live demo at TechCrunch Disrupt, apparently.”
Hicham Oudghiri Feb 21, 2016 ▶ 13:05
Prediction Not checkable as stated
Oudghiri: Software code will look more like data in the future
“And this, I tell you, will be a trend. Your code will look more like data than it does today.”
Hicham Oudghiri Feb 21, 2016 ▶ 14:07
Insight
Oudghiri: Data feedback loops are more valuable than underlying predictive models
“Some of the best companies I know have very strong feedback loops into their data, and that's, you know, an even better technology than the underlying model.”
Hicham Oudghiri Feb 21, 2016 ▶ 20:48
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.