Video summary

DSAI HDA AVS Technical Skill 1: Pengantar Data Analitics

Main summary

Key takeaways

Educational

Main Ideas and Lessons

Technical session context

  • This was a technical follow-up to a previous soft-skills session.
  • The audience is mainly semester 6 students from CS / Data Science / Cyber / Mobile tracks.
  • The speaker frames the topic as introductory / semi-technical, with more advanced topics planned later.

Why data analytics matters in the digital transformation era

  • Modern life generates continuous data (IoT, apps, transactions, streaming, smart devices).
  • This leads to rapid growth in data volume (“big data”).
  • Organizations need more than data—they need answers/insight to support decisions, such as:
    • greater clarity and confidence,
    • efficient allocation of resources.

What data and analytics mean

  • Data is presented as a collection of facts (raw material).
  • Data becomes useful when analyzed/processed into information that supports:
    • problem-solving,
    • decision-making.
  • Data analytics is described as interdisciplinary:
    • Statistics (analysis perspective),
    • Computing / IT (processing at scale),
    • Art (visualization and storytelling).

“Ocean of data” → clarity through a pipeline

  • Raw data is compared to unprocessed ingredients (e.g., avocado).
  • Analytics is the process that turns it into something valuable (e.g., juice).
  • Core promise: transform raw/huge data into accurate strategic insights that enable action.

Data Analytics Pipeline (Conceptual Methodology)

The workflow moves from data to action as a cycle:

  1. Data checking / data quality inspection

    • Use a “garbage in, garbage out” mindset to remove/handle problematic data.
    • Common bad data examples:
      • Redundancy: duplicate datasets (e.g., overlapping government population data)
      • Inconsistency: ambiguous coding (e.g., unclear gender encoding)
      • Outliers/anomalies:
        • treated differently depending on domain

  2. Transformation / data preparation

    • Convert raw data into a format ready for processing (like peeling/cutting ingredients).

  3. Modeling

    • Apply statistics / ML / AI to find patterns and predict outcomes.

  4. Decision-making + action

    • Outputs support real operational decisions.

  5. Feedback loop

    • Actions and outcomes feed back into the next round of processing/analysis.

Outliers: “Bad Data” vs “Valuable Signals”

  • In many analyses, outliers can be treated as errors or anomalies to clean.
  • In cybersecurity / fraud detection, anomalies can be the most important signals.

Examples:

  • Unusual bank transactions at odd hours → potential fraud
  • IoT sensor alerts (e.g., engine temperature outside limits) → early warning of failure/defects
  • Early warning systems and stress testing using extreme scenarios

Four Pillars of Data Analytics (Maturity Levels)

  1. Descriptive analytics — “What happened?”

    • Summarizes current/past performance using reports/dashboards.

  2. Diagnostic analytics — “Why did it happen?”

    • Finds cause-and-effect relationships (more “detective work”).

  3. Predictive analytics — “What will happen next?”

    • Forecasts trends/risks using historical data + ML/statistics.
    • Example: factory sensors predicting breakdown/defects days ahead.

  4. Prescriptive analytics — “What should we do?”

    • Recommends actions and can enable decision automation.
    • Includes:
      • Decision support: system recommends; humans choose
      • Decision automation: system takes actions automatically
  • The speaker uses a factory scenario to show progression:
    • reports → warnings → automated actions.

Business Analytics vs Data Analytics

  • Business analytics builds on data analytics by adding business context and goals.
  • It follows an analytics lifecycle (below).

Analytics Lifecycle for Business Analytics (Step-by-Step)

  1. Discovery

    • Clarify business context, expected outcomes, and relevant problems/events.
    • Identify available data/assets.

  2. Data preparation

    • Prepare datasets needed for modeling.

  3. Planning & building model

    • IT teams/analysts design algorithms and develop the model.

  4. Communicate results

    • Share insights with stakeholders and collect feedback.

  5. Operationalize (implementation)

    • Integrate analytics into operations and loop feedback back into discovery.

Expected outcomes:

  • smarter decisions,
  • better customer experience,
  • operational efficiency,
  • innovation.

Tooling Toolkit (How to Practice Analytics in Real Work)

The speaker groups skills/tools into three levels:

  1. Foundational processing

    • Example tools: Microsoft Excel, SQL (e.g., MySQL)

  2. Statistical computing / advanced modeling

    • Example tools/languages: Python, R
    • Encourages experimenting with AI models (including generative AI).
    • Notes that AI-assisted coding can be an option.

  3. Professional visualization / storytelling

    • Example tools: Tableau, Power BI, Lucid/Similar, ELK/Kibana (mentioned)
    • Emphasizes:
      • choosing the right charts,
      • real-time monitoring,
      • persuasive narrative.

Narrative and Stakeholder Communication

  • A key skill is translating numbers into stakeholder-appropriate storytelling.
  • Presentations should be:
    • persuasive,
    • understandable,
    • tailored to the audience.

Career-Role Clarification and Expectations

In Q&A, roles were clarified (noting titles can overlap in real organizations):

  • Data analysts
    • more stakeholder presentation and analysis
  • Data scientists
    • stronger focus on statistical modeling / ML/statistics foundations
  • Data engineers
    • build and maintain the data infrastructure

Future Trends and Concerns

  • Long-term advantage: being able to understand data (attributed to “Davenport” research).
  • Increasing investment in data infrastructure.
  • More emphasis on governance and ethics, especially with AI.
  • Explainable AI (XAI) focus, especially in health contexts.
    • Example mentioned: “Kemenkes 1 Sehat” initiative
    • Point: medical AI needs explanation/justification, not only outputs.
  • Continued professional growth and strong demand for roles like data engineers.

Methodologies / Instructions

A) “How data analytics turns raw data into value” (Conceptual workflow)

  • Start from data (facts/raw material).
  • Inspect and validate data (garbage-in/garbage-out):
    • handle redundancy
    • resolve inconsistency
    • decide how to treat outliers
      • clean/remove in general analysis
      • treat anomalies as signals in security/fraud
  • Transform into processing-ready format.
  • Model (statistics / ML / AI) to find patterns and learn.
  • Produce:
    • insights → enable decisions → lead to actions
  • Loop: actions → outcomes → updated data/analysis.

B) Four-steps maturity model: descriptive → diagnostic → predictive → prescriptive

  • Descriptive: what happened?
  • Diagnostic: why did it happen?
  • Predictive: what will happen next?
  • Prescriptive: what should we do? (recommend or automate actions)

C) Business analytics lifecycle

  • Discovery → Data preparation → Planning & building model
  • Communicate + operationalize
  • Feedback loop back to discovery/planning based on real outcomes.

D) Practical “tool usage” guidance (what to learn first)

  • Don’t try to master every tool at once.
  • Choose an entry point:
    • prefer foundations → Excel / SQL
    • prefer modeling → Python / R and ML/AI experiments
    • prefer communication → visualization tools and storytelling
  • The goal is clarity → decision → action, not only “pretty graphs.”

Speakers / Sources Featured

Speakers

  • Mr. Adit Handita Kusuma
    • Professional IT consultant
    • referenced as General Director/CEO of:
      • Tigos Tekno Teknoreif
      • Tugu Caraka Mitra Sinergi
  • Session host/organizers/participants (mentioned):
    • Ms. Jurige / Mrs. Jurik / Ms. Jeslin / Ms. Ajeng
    • Ms. Jingaini / Miss Jingaini
    • Ms. (operator/admin) (Teams/screen-sharing/evaluation guidance)
  • Students / participants in Q&A:
    • Jonathan
    • Mas Alvin
    • Mas Aloisius Probos Suryo / Aloisius
    • Mas Alvin / Alvin (mentioned again as a “young CEO candidate” in explanation)

Sources / referenced research/entities

  • IDC, Statistak, Domo, Cisco (big data production stats)
  • “Devenport” / Davenport (research concept: winning organizations understand data)
  • Kemenkes 1 Sehat (healthcare AI/XAI context)
  • DPR and Indonesian government/Ministry references (e-KTP copying case mentioned)
  • Philip Morris (factory/production example used in scenarios)
  • Japan tsunami early warning system (early warning analogy)
  • LinkedIn (career shift discussions into IT)

Original video