Ground more wiki pages in podcast evidence

Author: alexeygrigorev

CONTENT_TODO.md

@@ -14,10 +14,10 @@ Follow these rules when adding any page from this backlog.
   selected.
 - Keep `_podcast_summaries/` compact enough for agents to decide whether they
   need the source episode.
-- Link podcast evidence to specific episode URLs such as
-  `https://datatalks.club/podcast/<source-file-slug>.html` when the source
-  episode is known. Use `https://datatalks.club/podcast.html` only as a
-  temporary fallback.
+- Link podcast evidence to local podcast summary pages such as
+  `{{ '/podcasts/<source-file-slug>/' | relative_url }}`. Those local pages
+  link to the canonical DataTalks.Club episode. Do not link article or wiki
+  evidence to the generic podcast archive page.
 - Run `make check` after adding pages so graph and search stay current.
 
 ## Current Rewrite Requirements

_articles/apache-airflow.md

@@ -1,84 +1,171 @@
 ---
 layout: wiki
 title: "Applied Research"
-summary: "How the archive describes applied research as hypothesis-driven ML work that turns uncertain ideas into reusable artifacts, products, and production systems."
+summary: "How DataTalks.Club guests describe applied research as hypothesis-driven work that turns uncertain ML ideas into products, reusable systems, and production-ready evidence."
 related:
   - Machine Learning
   - Production
   - Machine Learning System Design
   - Academic Researcher to Data Science
-  - AI
+  - Experimentation
+  - MLOps
 ---
 
-## Definition and Scope
-
-Teams do applied research when they aim research at a practical product,
-workflow, or business decision. In the podcast archive, applied researchers
-produce datasets and models. They also produce benchmarks, experiments,
-prototypes, and product insights. The work is still scientific, but it's
-anchored to a use case and domain. Production constraints guide it too.
-
-Use this page when evidence connects research to practical ML systems. Use
-[Academic Researcher to Data Science]({{ '/wiki/academic-researcher-to-data-science/' | relative_url }})
-for career-transition material.
-
-
-## Recurring Archive Themes
-
-Teams start applied research with hypotheses, not only models. Vin Vashishta
-describes the first model as a hypothesis about how a system works. The
-researcher then uses data, experiments, model behavior, and explainability to
-refine that hypothesis. Production feedback can confirm or reject it.
-
-Research teams should make artifacts reusable. Guests treat novel datasets,
-benchmarks, model prototypes, and experimental results as intellectual property
-that downstream data scientists and ML engineers can reuse. Reusable research is
-valuable when it improves the next product decision or production build.
-
-The research-to-production gap is organizational as much as technical. Mihail
-Eric argues for embedded teams, role fluidity, researcher code reviews, and
-engineers who read papers or reproduce models. Researchers and engineers
-struggle when a team throws notebooks over the wall with unclear assumptions.
-
-When teams deploy research in the real world, the research question changes.
-Computer-vision and autonomous-driving episodes move from model quality to
-latency and sensor cost. They also bring in labeling, simulation, staged
-rollouts, and privacy. Safety checks, edge cases, and system coordination
-become part of the work too.
-
-## Episode Evidence
-
-These episodes give the strongest applied-research evidence.
-
-- [Monetize Machine Learning](https://datatalks.club/podcast.html)
-  (20:15-29:18) separates applied researchers from science
-  researchers and frames applied research as business-aligned artifact creation.
-- [Monetize Machine Learning](https://datatalks.club/podcast.html)
-  (36:10-43:28) defines research as hypothesis design,
-  experimentation, and explainability while also covering model deconstruction
-  and iterative evidence gathering.
-- [From Research to Production](https://datatalks.club/podcast.html)
-  (8:34-20:25) covers research infrastructure, prototyping, and benchmarks while
-  Mihail also discusses experimental tooling and deployment stacks.
-- [From Research to Production](https://datatalks.club/podcast.html)
-  (23:32-51:28) covers reproducibility, engineering rigor,
-  embedded teams, and code reviews. Later sections cover end-to-end systems,
-  paper reading, and model reproduction.
-- [Applying Computer Vision Research to Production AI](https://datatalks.club/podcast.html)
-  (29:45-36:12) describes simulation, closed-track testing, and
-  on-road testing. Aishwarya also covers anonymized sensor data, labeling
-  strategy, safety checks, and staged deployments.
-- [Applying Computer Vision Research to Production AI](https://datatalks.club/podcast.html)
-  (51:28-55:25) discusses sensitive-case evaluation, broader scenario
-  tests, gradual rollout, and multimodal LLM constraints.
+## Definition
+
+Teams do applied research when they aim research at a practical decision,
+product, or system. In the DataTalks.Club archive, guests use the term for work
+that still has scientific uncertainty but must produce something a team can use.
+That output may be a dataset, benchmark, or prototype. It may also be a
+modeling approach, validation method, or production design.
+
+That puts applied research between [machine learning]({{ '/wiki/machine-learning/' | relative_url }}),
+[experimentation]({{ '/wiki/experimentation/' | relative_url }}), and
+[production]({{ '/wiki/production/' | relative_url }}). It isn't only paper
+reading or model training. It also includes the work needed to decide whether an
+idea can survive product constraints, engineering constraints, and real users.
+
+## Common Definition
+
+Across the interviews applied research means hypothesis-driven technical work
+with a use case attached. In
+[Monetize Machine Learning]({{ '/podcasts/make-money-with-machine-learning-roles-skills/' | relative_url }}),
+[Vin Vashishta]({{ '/people/vinvashishta/' | relative_url }}) describes machine
+learning research as artifact creation. Teams create datasets, run experiments,
+study model behavior, and build knowledge that can support ML products. Around
+the 36:10 chapter, he frames the research skill set around hypothesis design and
+experimentation. He also includes explainability and iterative evidence
+gathering.
+
+[Mihail Eric]({{ '/people/mihaileric/' | relative_url }}) gives the matching
+engineering version in
+[From Research to Production]({{ '/podcasts/research-to-production-ml-systems-roadmap/' | relative_url }}).
+At 8:34 and 10:52, he connects research infrastructure with data collection and
+prototyping. He also connects it with hypotheses and benchmarks. His version of
+applied research doesn't stop when a notebook works. Researchers continue until
+the idea can move
+toward a reproducible [ML system design]({{ '/wiki/machine-learning-system-design/' | relative_url }}).
+
+The common definition is practical. A team tests an uncertain technical idea and
+turns what it learns into reusable evidence for product, engineering, or
+business decisions.
+
+## Guest Differences
+
+Guests agree on the hypothesis-driven part, but they focus on different outputs.
+
+[Vin Vashishta]({{ '/people/vinvashishta/' | relative_url }}) puts applied
+research close to [data products]({{ '/wiki/data-products/' | relative_url }})
+and ML monetization. In
+[Monetize Machine Learning]({{ '/podcasts/make-money-with-machine-learning-roles-skills/' | relative_url }}),
+the 20:15 and 26:58 chapters separate research work from pure product
+management and pure architecture. Researchers create the technical evidence
+that lets a team decide whether a model can become a revenue-generating product.
+
+[Mihail Eric]({{ '/people/mihaileric/' | relative_url }}) puts the boundary
+between research and [MLOps]({{ '/wiki/mlops/' | relative_url }}) under
+pressure. In
+[From Research to Production]({{ '/podcasts/research-to-production-ml-systems-roadmap/' | relative_url }}),
+the 23:32, 30:16, and 34:20 chapters argue that researchers need engineering
+rigor and engineers need experimental rigor. He describes embedded teams and
+code reviews for researchers. He also describes engineers who read papers or
+reproduce models.
+
+[Aishwarya Jadhav]({{ '/people/aishwaryajadhav/' | relative_url }}) makes the
+definition domain-specific. In
+[Applying Computer Vision Research to Building Production-Ready AI Systems]({{ '/podcasts/from-computer-vision-research-to-autonomous-driving-ai/' | relative_url }}),
+she ties applied research to sensors, latency, and labeling strategy. She also
+ties it to safety checks and staged deployment. The 29:45 and 32:43 chapters
+connect [computer vision]({{ '/wiki/computer-vision/' | relative_url }})
+research in autonomous driving with simulation and closed-track tests. They also
+cover on-road tests and release controls before the system becomes usable.
+
+## Turning Research Into Product Decisions
+
+Teams benefit from applied research when it changes a decision. In Vin
+Vashishta's
+[Monetize Machine Learning]({{ '/podcasts/make-money-with-machine-learning-roles-skills/' | relative_url }}),
+the 43:28 chapter connects product management with researchable use cases. A
+business question has to become a technical hypothesis before researchers can
+test it. The 48:54 chapter adds feasibility studies and gated decisions, which
+keeps research connected to product risk instead of open-ended exploration.
+
+[Loris Marini]({{ '/people/lorismarini/' | relative_url }}) describes a similar
+SaaS case. In
+[Data Professionals Need Business Skills in SaaS]({{ '/podcasts/data-professionals-business-skills-in-saas/' | relative_url }}),
+the 2:45 chapter describes research skills applied in a startup. The 8:30
+chapter discusses reinforcement learning applied to practical problems. That
+episode ties applied research to stakeholder language and business metrics. It
+also ties the work to model deployment rather than treating research as a
+separate track.
+
+For a data team, applied research should produce a decision-ready answer. The
+team should know whether to continue, stop, or simplify. It should also know
+whether to collect different data, change the metric, or move toward production.
+That connects the topic to
+[experimentation and causal inference]({{ '/wiki/experimentation-and-causal-inference/' | relative_url }})
+when the main question is evidence quality. It connects to
+[machine learning system design]({{ '/wiki/machine-learning-system-design/' | relative_url }})
+when the main question is whether the approach can be served reliably.
+
+## Engineering And Reproducibility
+
+The archive repeatedly treats reproducibility as part of applied research, not
+as cleanup after the fact. In
+[From Research to Production]({{ '/podcasts/research-to-production-ml-systems-roadmap/' | relative_url }}),
+[Mihail Eric]({{ '/people/mihaileric/' | relative_url }}) says researchers need
+engineering rigor at the 23:32 chapter. In the 44:36 and 46:57 chapters, he
+recommends end-to-end systems, deployment practice, and code reviews. If no one
+can reproduce the result or run the system, the research can't guide production
+work.
+
+[Johanna Bayer]({{ '/people/johannabayer/' | relative_url }}) makes the same
+argument from academia in
+[Teaching Open Science and Reproducible Research]({{ '/podcasts/teaching-reproducible-research-and-open-science-coding-practices-for-academia/' | relative_url }}).
+At 8:30, she discusses reproducible manuscripts with embedded code. At 12:10
+and 16:36, she frames research software engineering around software-focused
+research outputs and reusable toolboxes. Her
+discussion connects applied research with [open source]({{ '/wiki/open-source/' | relative_url }})
+and [software engineering]({{ '/wiki/software-engineering/' | relative_url }}).
+
+[Eleni Tzirita Zacharatou]({{ '/people/elenitziritazacharatou/' | relative_url }})
+adds a systems-research version in
+[Big Data Analytics and Postdoc Research]({{ '/podcasts/big-data-analytics-and-postdoc-research/' | relative_url }}).
+The 23:08 and 24:15 chapters discuss Nebula Stream and Agora infrastructure.
+They also discuss system-driven research that grows from earlier
+stream-processing systems such as Apache Flink. In that setting, researchers do
+not only produce an insight. They also produce an infrastructure idea that other
+researchers or industry teams can
+evaluate.
+
+## Domain Constraints
+
+The work changes when the deployment domain changes. In
+[Applying Computer Vision Research to Building Production-Ready AI Systems]({{ '/podcasts/from-computer-vision-research-to-autonomous-driving-ai/' | relative_url }}),
+[Aishwarya Jadhav]({{ '/people/aishwaryajadhav/' | relative_url }}) discusses
+sensor tradeoffs at 11:22 and on-vehicle inference at 22:17. She covers model
+compression at 23:28 and staged validation from 29:45 onward. Those constraints
+make the research question more specific than "which model is best?" The team
+has to ask which model works under the product's cost, latency, privacy, and
+safety requirements.
+
+In the same episode, the 51:28 chapter discusses sensitive-case testing and the
+52:53 chapter discusses multimodal LLMs in autonomous-driving contexts. This is
+why applied research often links to [LLM production patterns]({{ '/wiki/llm-production-patterns/' | relative_url }})
+and [notebook-to-production AI systems]({{ '/wiki/notebook-to-production-ai-systems/' | relative_url }}).
+The research idea has to meet evaluation, rollout, monitoring, and system
+coordination constraints before a team can operate it.
 
 ## Related Pages
 
-Use these pages for adjacent ML, production, and career material.
+These pages cover the adjacent roles, systems, and practices that appear in the
+same podcast discussions.
 
 - [Machine Learning]({{ '/wiki/machine-learning/' | relative_url }})
+- [Experimentation]({{ '/wiki/experimentation/' | relative_url }})
+- [MLOps]({{ '/wiki/mlops/' | relative_url }})
 - [Production]({{ '/wiki/production/' | relative_url }})
 - [Machine Learning System Design]({{ '/wiki/machine-learning-system-design/' | relative_url }})
 - [Academic Researcher to Data Science]({{ '/wiki/academic-researcher-to-data-science/' | relative_url }})
-- [AI]({{ '/wiki/ai/' | relative_url }})
-- [MLOps]({{ '/wiki/mlops/' | relative_url }})
+- [Computer Vision]({{ '/wiki/computer-vision/' | relative_url }})
+- [Open Source]({{ '/wiki/open-source/' | relative_url }})