CV β†“

PORTFOLIO / NUR FAJAR β€” TANGERANG, ID

Building AI systems.
Training the people who use them.

  • Open to work
  • Relocation OK
  • Remote OK

Learning & Development Specialist β€” most recently at a Singapore-based AI company, where I designed GenAI curriculum end-to-end and engineered the automation pipelines behind it. Two tracks, one signal.

350+Learners trained
9.0/10Satisfaction
9AI agents deployed
3.94GPA Β· best graduate

Learners cumulative since 2023 Β· satisfaction is a Kirkpatrick L1 mean

Portrait of Nur Fajar
00:00

Your move

TRACKS

Track 01 / AI SYSTEMS

AI engineering & automation

Engineered an end-to-end B2B outreach automation pipeline in Python β€” 9 specialized AI agents across the CRM covering lead qualification, company research, cold email drafting, follow-ups, and inbox monitoring, with human-in-the-loop approval before anything sends. Live Jun–Jul 2026: ~3,000 leads processed and 600 qualified prospects surfaced, at under 5 minutes of prep per prospect instead of ~30. By hand that volume is roughly 1,250 hours of work β€” which is why it was automated rather than staffed.

  • Python
  • LLM APIs
  • AI agents
  • CRM automation
  • HITL

Track 02 / LEARNING & DEVELOPMENT

GenAI curriculum & delivery

Designed 4 GenAI training modules β€” business, education, developer, and product management tracks β€” and led 5 programs end-to-end, from learner analysis and backward design to live delivery and Kirkpatrick-style evaluation. Training and mentoring since 2023 β€” 350+ learners cumulative across public, corporate, and institutional cohorts, at a 9.0/10 mean satisfaction score (Kirkpatrick L1), plus a Train the Trainers initiative that upskilled 6 university lecturers.

  • Instructional design
  • ADDIE
  • Backward design
  • Facilitation
  • TTT

REFERENCES / WHAT PEOPLE SAY

REF / 01 Β· MANAGER

I managed Nur Fajar directly at Terra AI. He mentored more than 100 students and consistently earned outstanding satisfaction scores from them β€” 9.8 out of 10. What stood out to me most was his organizational skill and his initiative: he took it upon himself to build a chatbot that helped students track their own learning progress. I'd recommend him without hesitation.
Christian Jonathan Manager Β· Terra AI

REF / 02 Β· L&D PRACTITIONER

He led with real skill and deep knowledge, and he never let a mentee's progress go unnoticed β€” he made a point of supporting us and recognizing every milestone along the way.
Kevin Naufal Eryogia Mentee, Bangkit 2023 Β· now People Team Coordinator, MondelΔ“z International

REF / 03 Β· MENTEE

His technical depth came paired with real leadership and mentoring ability β€” he had a way of making complex ideas accessible no matter where you were starting from. That versatility would serve him well as a Project Manager, Program Manager, in HR, or in Learning & Development.
Andrew Benedictus Jamesie Mentee, Bangkit 2023 Β· now Senior Software Engineer, Accenture

REF / 04 Β· TEAMMATE

He ran the team with real skill β€” communication stayed clear, tasks were managed effectively, and the whole group worked well together right through to a successful finish.
Diki Hamdani Capstone teammate Β· Bangkit Academy 2023

REF / 05 Β· COLLEAGUE

His communication was excellent β€” always clear and efficient. His management stood out too: organized, proactive, and positive, with everything delivered on time.
Dicky Arya Pratama Colleague, GDSC Indonesia Β· Senior Android Developer

Six further recommendations, including from mentees now at Apple Developer Academy alumni programmes and Accenture, are on LinkedIn.

SIGNALS / WORK EXPERIENCE

EXP / 01 Learning & Development Specialist Terra Weather Pte. Ltd. Β· Singapore (Remote) Β· Jul 2025 – Jul 2026
  • Developed 4 GenAI training modules spanning business, education, and product management tracks β€” foundational to production-ready level.
  • Led 5 learning programs end-to-end, reaching 150+ participants across webinars and workshops; upskilled 6 university lecturers via Train the Trainers.
  • Engineered a B2B outreach automation pipeline in Python with LLM APIs; ran live Jun–Jul 2026 β€” ~3,000 leads processed, 600 qualified prospects surfaced.
  • Deployed 9 specialized AI agents across the CRM pipeline with human-in-the-loop approval before sending.
  • Cut per-prospect email prep from ~30 min to under 5 (βˆ’83%) β€” at ~3,000 leads, a manual-equivalent workload of ~1,250 hours that never had to be staffed.
  • Produced 4 alumni testimonial videos from 30 outreach contacts and 8 interviews, supporting learner acquisition.
EXP / 02 AI Training Specialist Terra AI Β· Singapore (Remote) Β· Feb 2024 – Jun 2025
  • Trained 100+ students and professionals on generative AI, prompt engineering, and chatbot development β€” customized curriculum for startups through multinationals.
  • Sustained a 9.0/10 mean learner satisfaction score (Kirkpatrick L1) across all cohorts via structured feedback loops and real-time curriculum iteration.
  • Ran a 20% instruction / 80% practice delivery framework with project-based assessments and learner progress dashboards.
EXP / 03 Machine Learning Mentor Bangkit Academy Β· Ministry of Education program Β· Feb 2023 – Jan 2024
  • Mentored 50+ students from 25+ universities across Indonesia; cohort graduation rate above 90%.
  • Conducted 40+ weekly sessions covering technical skills, soft skills, and engagement activities.
  • Liaised between participants and the Bangkit team, coordinating with 20+ industry and academic experts.

ORGANIZATIONAL EXPERIENCE

Enablement work is facilitation work, and facilitation takes reps. These are the three that built them β€” a chapter founded from zero, a 2,000-person programme, and a village team of 16. Eight organisations in total; the rest are on LinkedIn.

ORG / 01 Tech Community Chapter Lead Google Developer Student Clubs (GDSC) Unsil Β· Aug 2021 – Jul 2022
  • Founded and led the first-generation GDSC chapter at Universitas Siliwangi, recruiting 100 new members across cohorts in the first year.
  • Planned and ran a full annual program slate β€” Android Study Jam, Flutter Festival, a hackathon, and a Career Talk Series β€” mostly in collaboration with other GDSC chapters across Indonesia.
  • Output: 100 new members recruited in year one; event series averaged 50+ attendees per session.
ORG / 02 Programme Delivery at Scale Lead Organizer, Kuliah Dhuha 2021 LDK KISI Β· Mentoring Division Β· 2021
  • Lead organizer of a large-scale programme welcoming new students through an Islamic mentoring lens, directing a cross-divisional team of about 50 people.
  • Output: 2,000+ attendees online; organizing team of ~50 across divisions.
ORG / 03 Community Leadership Village Coordinator (Kordes) KKN Β· Purwaharja Village, Banjar City Β· Feb–Mar 2021
  • Led a cross-disciplinary community service group of 16 students for one month, designing and executing 10 work programmes β€” including a Google Workspace workshop for local schools and village staff, and tutoring for school-age children.
  • Output: 10 programmes completed; ~30 workshop participants, ~40 tutoring participants, ~200 total direct beneficiaries.

SKILL GROUPS

SKL / 01 β€” AI ENGINEERING

  • Python
  • LLM APIs (GPT-4o-mini)
  • Prompt engineering
  • AI agents
  • CRM automation
  • Email automation
  • HITL pipelines
  • Data enrichment
  • TensorFlow
  • Sentiment analysis
  • Ensemble learning
  • SMOTE

SKL / 02 β€” LEARNING & DEVELOPMENT

  • Instructional design
  • Curriculum development
  • ADDIE
  • Backward design
  • Bloom's taxonomy
  • Kirkpatrick L1–L3
  • Train the Trainers
  • Facilitation
  • Cohort-based learning
  • Learner analytics
  • LMS management
  • Assessment design

SKL / 03 β€” PRODUCT & PROGRAM

  • Program management
  • Design thinking
  • Jobs-to-be-Done
  • MoSCoW prioritization
  • Stakeholder mapping
  • Community building

SKL / 04 β€” TOOLS

  • Google Workspace
  • Notion
  • Miro
  • Discord
  • Smojo
  • Apollo.io
  • Adobe Premiere Pro

PROGRAMS / CURRICULUM DESIGNED

Four published LMS courses, 20 hours of contact time. Every course is backward designed from a published artifact: the learner does not leave with notes, they leave with something running. Open a card for the unit breakdown.

PRG / 01 Β· BEGINNER Β· 2H

Chatbots for Education

Non-technical teachers ship a published, AI-integrated educational chatbot in 2 hours β€” design thinking first, template-based build second.

Audience
School teachers Β· Lecturers
Prerequisites
None
Unit breakdown
  1. Problem FramingEducator pain points Β· AI chatbot benefits Β· real-world examples from prior cohorts
  2. Design Thinking WorkshopSimplified 4-step framework β€” Why / For whom / About what / How Β· design worksheet completion
  3. Build & DeployTemplate customization Β· content authoring Β· ChatGPT integration Β· publish and test

Learner leaves with: Design Thinking worksheet Β· published educational chatbot with design rationale

Open on ai4impact β†—

PRG / 02 Β· BEGINNER Β· 2H

Chatbots for Business

Business owners deploy a live, branded customer-service chatbot in under 15 minutes, then customize and harden it against hallucination.

Audience
SMB owners Β· Managers
Prerequisites
None
Unit breakdown
  1. Business Case + First LaunchROI framing Β· chatbot vs. website FAQ Β· live chatbot in under 15 min
  2. Customization & BrandIndustry-specific FAQs Β· tone and persona Β· welcome flow design
  3. AI Integration + RiskLLM API activation Β· hallucination & scope creep Β· responsible deployment

Learner leaves with: branded business chatbot on a live URL Β· failure-mode diagnosis skills

Open on ai4impact β†—

PRG / 03 Β· INTERMEDIATE Β· 4H

GenAI Foundations

Developers build structured mental models of LLMs β€” from NLP history to transformer mechanics β€” through a progressive lab series ending in a deployed hybrid app.

Audience
Backend & full-stack engineers
Prerequisites
Programming experience in at least one language
Unit breakdown
  1. NLP History & LLM OriginsTemplate matching Β· grammar parsing Β· why both failed Β· what LLMs actually solve
  2. LLM FundamentalsTransformer architecture Β· tokenization Β· context window Β· inference mechanics
  3. Lab Series: Build & DeployEnvironment setup Β· progressive code builds Β· API integration Β· deployment and testing

Learner leaves with: deployed LLM application Β· API integration skills Β· a working mental model of LLM architecture

Open on ai4impact β†—

PRG / 04 Β· SPECIALIST Β· 12H

GenAI Product Manager

Mid-to-senior PMs learn to own AI decisions β€” rule-vs-GPT trade-offs, risk thresholds, and prompt design as product decisions β€” via a real chatbot pivot case study.

Audience
Mid-to-senior PMs Β· Product leads
Prerequisites
JTBD, product discovery, and roadmapping experience
Unit breakdown
  1. PM Mindset & Product ContextDecision ownership under AI uncertainty Β· gap analysis Β· Keep/Change/Remove Β· reuse vs. rebuild
  2. User Definition & Product IdentityPersona + JTBD for AI products Β· UI/UX as PM decisions Β· feature scope Β· MoSCoW prioritization
  3. Product Strategy & AI IntegrationValue proposition design Β· rule-based vs. GPT trade-off Β· prompt engineering as a PM decision Β· analytics-driven iteration

Learner leaves with: published chatbot plus a full PM decision log Β· product gap analysis Β· AI feature prioritization skills

Open on ai4impact β†—

Design frameworks applied across the set: backward design and Bloom's taxonomy in all four Β· design thinking in both chatbot courses Β· JTBD and MoSCoW in the PM track Β· problem-based learning in Foundations.

INSTRUCTIONAL DESIGN / CASE STUDY

One course taken apart, so the design reasoning is visible rather than asserted. Chatbots for Business β€” 2 hours, 3 lessons, non-technical business owners, and a hard constraint: the session must end with a live published product, not a demo.

01 Β· DESIGN CONTEXT

Problem
SMBs lose potential customers to unanswered inquiries β€” particularly outside business hours β€” but have no technical team to build automated solutions.
Gap
Learners need to deploy a branded, functional AI chatbot. The barrier is not motivation, it is technical confidence.
Constraint
A 2-hour synchronous session that must end with a live, published product.

02 Β· LEARNER PROFILE

Prior knowledge
No programming background. Comfortable with spreadsheets and basic web tools.
Motivation
ROI-driven. Wants to cut repetitive customer questions and improve response time without hiring.
Fears
Breaking something technical they cannot fix Β· investing time in a tool that will not work for their business Β· looking incompetent in front of customers.
Design implication
Confidence-building must precede complexity. Early success is a prerequisite for sustained engagement, not a bonus.

03 Β· LEARNING OBJECTIVES Bloom-tagged, backward designed from the artifact

  • Understand Articulate the business case for AI chatbots over static FAQ pages, including response time, availability, and cost trade-offs.
  • Apply Customize and deploy a branded business chatbot from a template inside a live session environment.
  • Apply Integrate an LLM API to handle open-ended customer queries that fall outside scripted flows.
  • Evaluate Identify and mitigate common chatbot failure modes: hallucination, scope creep, and brand inconsistency.

04 Β· SESSION ARCHITECTURE

  1. Lesson 1 β€” Business Case + First Launch 35 min

    Goal: learner has a live chatbot before the lesson ends

    • Hook Dental clinic scenario β€” a customer inquiry arrives at 11 PM on a Sunday. What happens next?
    • Concept Chatbot vs. website FAQ: availability, response quality, lead capture, cost per query.
    • Activity Live build: open template β†’ rename business β†’ set welcome message β†’ publish. Target 12 minutes.
    • Reflection Share the live URL with a peer. First success checkpoint.

    Design note β€” "launch first, customize later": learners commit to the tool before they invest effort in it, which cuts the early dropout caused by abstract setup tasks.

  2. Lesson 2 β€” Customization & Brand Identity 30 min

    Goal: the chatbot reflects the learner's real business context

    • Concept Brand elements in chatbot design: tone, persona, FAQ logic, welcome flow.
    • Activity Systematic customization β€” business name, at least 5 industry-specific FAQs, interaction tone, visual identity.
    • Check Peer test: does this chatbot actually answer your real customer questions?

    Design note β€” real business context removes abstraction. Learners are solving their own problem, not a hypothetical one.

  3. Lesson 3 β€” AI Integration + Risk Management 35 min

    Goal: LLM integrated, and the learner can name and handle failure modes

    • Concept Why LLMs: handling unpredictable questions that scripts cannot anticipate.
    • Activity Activate the LLM API in the existing chatbot β†’ test with unexpected queries β†’ observe behavior.
    • Concept Failure modes: hallucination (confidently wrong), scope creep (answering out-of-domain), brand drift.
    • Activity Diagnosis exercise β€” observe 3 chatbot behaviors, identify the failure type, apply the fix.

    Design note β€” risk comes after the first success. Introducing failure modes before a learner has experienced value creates anxiety that blocks learning.

05 Β· ASSESSMENT

Two instruments, both testing whether the learner can do the thing rather than recall it β€” and neither asking the learner to rate their own competence.

  1. During each lesson

    Formative β€” behaviour-verification quizzes Apply / Evaluate

    Learners observe their own live chatbot, diagnose a described failure, then select and apply a fix. Tests application, not recall.

  2. End of session

    Summative β€” published artifact plus design rationale Create / Evaluate

    A live URL and three sentences explaining one design decision the learner made, and why. The rationale is what prevents rote completion β€” it is possible to finish the build without understanding it, but not to justify it.

EDUCATION & CREDENTIALS

EDU / 01

Siliwangi University

B.CS in Informatics Β· 2018 – 2022

GPA 3.94 β€” Best Graduate, Faculty of Engineering. Bank Indonesia & BRI scholarships. Thesis published in JOIV: ensemble ML + SMOTE for SDG sentiment analysis.

CRT / 01–03

Certifications

TensorFlow Developer Certificate β€” Google, 2024.
Google Data Analytics Certificate β€” 2023.
Top 100 Next Digital Talent β€” IndonesiaNext by Telkomsel (out of 6,000 applicants); MOS PowerPoint & BNSP Digital Marketing certifications.