Prompt Engineering
Designing, testing and refining structured prompts to improve AI output quality, consistency, relevance and instruction following.
Hello, I'm
I evaluate, improve and build AI-powered content, workflows and digital experiences. Combining prompt engineering, AI evaluation, generative AI and web development to turn ideas into practical, deployable digital solutions.
I combine artificial intelligence, digital creativity and web development to build practical solutions for real-world needs.
I'm an AI-focused digital professional based in Indonesia with hands-on experience in prompt engineering, generative AI, AI-assisted content production, response evaluation, web development and digital marketing.
My work focuses on using AI as a practical problem-solving tool — from designing structured prompts and evaluating AI-generated responses to creating websites, visual content and AI-assisted digital workflows.
I combine AI capabilities with HTML, CSS, JavaScript, design, marketing and rapid prototyping to transform ideas into usable and deployable digital products.
As a native Indonesian speaker, I can also contribute to AI projects involving Indonesian language evaluation, localization, cultural context and response quality assessment.
const professional = {
name:
"Aknal Hartanto",
location:
"Indonesia",
focus: [
"AI Training",
"AI Evaluation",
"Prompt Engineering",
"Generative AI",
"Web Development"
],
languages: [
"Indonesian",
"English"
],
remote:
true,
status:
"Open to opportunities"
};
Structured prompts, iterative prompting and AI output optimization.
Branding, marketing, social content and generative visual concepts.
AI-assisted development, rapid prototyping and responsive web experiences.
Exploring new AI workflows, tools, techniques and practical applications.
Currently exploring remote opportunities where I can contribute to AI training, evaluation, prompt engineering, localization, generative AI and digital production.
Discuss an opportunityPractical AI capabilities spanning prompt engineering, model response evaluation, Indonesian localization, generative content and AI-assisted development.
Designing, testing and refining structured prompts to improve AI output quality, consistency, relevance and instruction following.
Reviewing AI-generated responses for accuracy, relevance, instruction following, clarity, reasoning quality and overall usefulness.
Evaluating Indonesian-language AI outputs for naturalness, meaning, tone, cultural context and linguistic appropriateness.
Using generative AI to produce and iterate text, visual concepts, marketing assets, creative materials and digital content.
Combining AI-assisted development with HTML, CSS and JavaScript to rapidly prototype, build and deploy responsive digital experiences.
Structuring AI-assisted workflows for repetitive content, research, ideation, digital production and business tasks.
I don't treat AI output as automatically correct. My workflow combines clear instructions, iterative generation and human evaluation before an output is considered complete.
Define the objective, context, constraints and expected output.
Build structured instructions and provide relevant context.
Review accuracy, relevance, clarity and instruction following.
Iterate until the output meets the required quality standard.
A practical combination of generative AI, web technologies, deployment tools and digital production skills.
Selected projects demonstrating how I combine AI-assisted ideation, web development, design, content strategy and rapid deployment.
Transforming business concepts into responsive, deployable landing pages by combining human direction with AI-assisted ideation, copywriting, development and iterative refinement.
Small businesses and personal brands often need professional web presence without lengthy development cycles.
Translate business requirements into page architecture, content, visual direction and reusable development instructions.
Use generative AI for ideation, content, coding assistance and iteration while manually reviewing and refining the output.
Responsive HTML, CSS and JavaScript websites prepared for production deployment.
Real-world examples of responsive websites, business landing pages and digital experiences.
Corporate-style digital presence focused on communicating business identity, services and company information through a structured responsive interface.
Visit Live ProjectFood and beverage landing page designed to present menu offerings, brand identity and customer information in a simple mobile-friendly experience.
Visit Live ProjectDigital brand website exploring modern visual presentation, structured content and responsive web implementation.
Visit Live ProjectWeb service concept focused on landing page development, digital presentation and accessible website solutions for businesses and personal brands.
Visit Live ProjectDigital wedding experience designed around event information, visual storytelling and an accessible mobile-first interface.
Visit Live ProjectFood business website combining brand presentation, product information and customer-oriented digital communication in a responsive layout.
Visit Live ProjectAdditional practical applications of generative AI and prompt engineering.
Reusable prompt structures for writing, business tasks, ideation, marketing and structured AI outputs.
Generative visual workflows for branding, packaging, advertising, social media and concept development.
AI-assisted processes for developing, reviewing and refining long-form and short-form digital content.
Converting ideas into working web prototypes through structured AI-assisted development workflows.
AI accelerates the process. Human judgment determines the quality.
My approach is not simply to generate an output, but to review, refine and transform AI-generated material into something useful.
A practical demonstration of how I compare AI-generated responses using structured criteria such as accuracy, relevance, instruction following, language quality and cultural appropriateness.
This case study demonstrates my evaluation methodology and is not presented as a client or employer project.
"Jelaskan apa itu kecerdasan buatan kepada siswa SMP menggunakan Bahasa Indonesia yang sederhana. Gunakan maksimal tiga paragraf dan berikan satu contoh yang dekat dengan kehidupan sehari-hari."
Kecerdasan buatan atau Artificial Intelligence adalah teknologi yang memungkinkan sistem komputer melakukan berbagai tugas yang biasanya membutuhkan kecerdasan manusia.
Teknologi ini menggunakan algoritma, pemrosesan data dan berbagai metode komputasi untuk menghasilkan keputusan.
Contohnya adalah teknologi AI yang digunakan dalam berbagai aplikasi modern.
Generally correct, but the explanation remains abstract and the example is not sufficiently concrete for the intended audience.
Kecerdasan buatan atau AI adalah teknologi yang membuat komputer bisa melakukan beberapa tugas yang biasanya membutuhkan kemampuan manusia, seperti mengenali gambar, memahami pertanyaan, atau memberikan saran.
AI belajar dari banyak data dan pola. Karena itu, AI dapat membantu manusia menyelesaikan pekerjaan tertentu dengan lebih cepat, tetapi hasilnya tetap perlu diperiksa karena AI juga bisa membuat kesalahan.
Contoh yang dekat dengan kehidupan sehari-hari adalah rekomendasi video di aplikasi. Sistem mempelajari jenis video yang sering kita tonton lalu mencoba merekomendasikan video lain yang mungkin kita sukai.
Clearer language, concrete example, appropriate complexity and stronger adherence to the requested audience and format.
Both responses are evaluated using the same criteria to make the comparison systematic and explainable.
| Criterion | Response A | Response B | Evaluation |
|---|---|---|---|
| Accuracy Factual correctness | 8/10 | 9/10 | Both are broadly accurate, while Response B adds an appropriate limitation about possible AI errors. |
| Relevance Addresses user need | 6/10 | 10/10 | Response B is more closely aligned with the needs of a junior-high-school audience. |
| Instruction Following Follows explicit constraints | 8/10 | 10/10 | Both respect the paragraph limit, but Response B better satisfies the requirement for a relatable example. |
| Language Quality Clarity and naturalness | 6/10 | 9/10 | Response A uses more abstract terminology, while Response B is easier for younger readers. |
| Audience Fit Context and accessibility | 6/10 | 9/10 | The everyday recommendation-system example makes Response B more accessible to the requested audience. |
Response B is preferred because it follows the user's instructions more precisely, uses simpler and more natural Indonesian, provides a concrete everyday example and communicates the concept at an appropriate level for junior-high-school students.
It directly adapts the explanation to the requested student audience.
Technical concepts are expressed using simpler Indonesian without unnecessary terminology.
Video recommendations provide a familiar scenario that makes the concept easier to understand.
It correctly communicates that AI-generated results can still require human verification.
A good AI response is not simply one that sounds convincing. It should satisfy the user's actual request while remaining accurate, relevant and appropriately written.
Understand what the user actually wants.
Check whether explicit requirements are followed.
Identify unsupported, misleading or incorrect claims.
Remove unnecessary information and prioritize useful content.
Evaluate clarity, tone, grammar and naturalness.
Ensure the response suits the intended user and context.
I'm available for remote opportunities involving AI training, LLM evaluation, prompt engineering, Indonesian localization, generative AI, AI-assisted web development and digital production.
I'm open to discussing remote roles, freelance projects, contract opportunities and AI evaluation assignments.
From evaluating AI responses to building AI-assisted digital products, I'm interested in work where technology is applied thoughtfully to solve practical problems.
Let's Talk