Immediate engineering methods and finest practices: Be taught by doing with Anthropic’s Claude 3 on Amazon Bedrock

You may have probably already had the chance to work together with generative synthetic intelligence (AI) instruments (akin to digital assistants and chatbot functions) and seen that you simply don’t all the time get the reply you might be in search of, and that attaining it will not be easy. Giant language fashions (LLMs), the fashions behind the generative AI revolution, obtain directions on what to do, how you can do it, and a set of expectations for his or her response by way of a pure language textual content known as a immediate. The best way prompts are crafted tremendously impacts the outcomes generated by the LLM. Poorly written prompts will usually result in hallucinations, sub-optimal outcomes, and total poor high quality of the generated response, whereas good-quality prompts will steer the output of the LLM to the output we would like.

On this submit, we present how you can construct environment friendly prompts on your functions. We use the simplicity of Amazon Bedrock playgrounds and the state-of-the-art Anthropic’s Claude 3 household of fashions to exhibit how one can construct environment friendly prompts by making use of easy methods.

Immediate engineering

Immediate engineering is the method of rigorously designing the prompts or directions given to generative AI fashions to provide the specified outputs. Prompts act as guides that present context and set expectations for the AI. With well-engineered prompts, builders can benefit from LLMs to generate high-quality, related outputs. As an illustration, we use the next immediate to generate a picture with the Amazon Titan Image Generation mannequin:

An illustration of an individual speaking to a robotic. The particular person seems to be visibly confused as a result of he cannot instruct the robotic to do what he desires.

We get the next generated picture.

Let’s take a look at one other instance. All of the examples on this submit are run utilizing Claude 3 Haiku in an Amazon Bedrock playground. Though the prompts may be run utilizing any LLM, we talk about finest practices for the Claude 3 household of fashions. To be able to get entry to the Claude 3 Haiku LLM on Amazon Bedrock, consult with Model access.

We use the next immediate:

Claude 3 Haiku’s response:

The request immediate is definitely very ambiguous. 10 + 10 could have a number of legitimate solutions; on this case, Claude 3 Haiku, utilizing its inside data, decided that 10 + 10 is 20. Let’s change the immediate to get a distinct reply for a similar query:

1 + 1 is an addition
1 - 1 is a substraction
1 * 1 is multiplication
1 / 1 is a division

What's 10 + 10?

Claude 3 Haiku’s response:

10 + 10 is an addition. The reply is 20.

The response modified accordingly by specifying that 10 + 10 is an addition. Moreover, though we didn’t request it, the mannequin additionally supplied the results of the operation. Let’s see how, by way of a quite simple prompting approach, we will get hold of an much more succinct outcome:

1 + 1 is an addition
1 - 1 is a substraction
1 * 1 is multiplication
1 / 1 is a division

What's 10 + 10?

Reply solely as within the examples supplied and 
present no extra info.

Claude 3 Haiku response:

Properly-designed prompts can enhance consumer expertise by making AI responses extra coherent, correct, and helpful, thereby making generative AI functions extra environment friendly and efficient.

The Claude 3 mannequin household

The Claude 3 household is a set of LLMs developed by Anthropic. These fashions are constructed upon the most recent developments in pure language processing (NLP) and machine studying (ML), permitting them to know and generate human-like textual content with outstanding fluency and coherence. The household is comprised of three fashions: Haiku, Sonnet, and Opus.

Haiku is the quickest and most cost-effective mannequin in the marketplace. It’s a quick, compact mannequin for near-instant responsiveness. For the overwhelming majority of workloads, Sonnet is 2 occasions quicker than Claude 2 and Claude 2.1, with greater ranges of intelligence, and it strikes the best stability between intelligence and velocity—qualities particularly crucial for enterprise use instances. Opus is probably the most superior, succesful, state-of-the-art basis mannequin (FM) with deep reasoning, superior math, and coding skills, with top-level efficiency on extremely complicated duties.

Among the many key options of the mannequin’s household are:

• Imaginative and prescient capabilities – Claude 3 fashions have been educated to not solely perceive textual content but in addition photographs, charts, diagrams, and extra.
• Finest-in-class benchmarks – Claude 3 exceeds present fashions on standardized evaluations akin to math issues, programming workout routines, and scientific reasoning. Particularly, Opus outperforms its friends on many of the widespread analysis benchmarks for AI programs, together with undergraduate degree skilled data (MMLU), graduate degree skilled reasoning (GPQA), fundamental arithmetic (GSM8K), and extra. It displays excessive ranges of comprehension and fluency on complicated duties, main the frontier of basic intelligence.
• Diminished hallucination – Claude 3 fashions mitigate hallucination by way of constitutional AI methods that present transparency into the mannequin’s reasoning, in addition to improved accuracy. Claude 3 Opus exhibits an estimated twofold acquire in accuracy over Claude 2.1 on troublesome open-ended questions, lowering the chance of defective responses.
• Lengthy context window – Claude 3 fashions excel at real-world retrieval duties with a 200,000-token context window, the equal of 500 pages of knowledge.

To be taught extra concerning the Claude 3 household, see Unlocking Innovation: AWS and Anthropic push the boundaries of generative AI together, Anthropic’s Claude 3 Sonnet foundation model is now available in Amazon Bedrock, and Anthropic’s Claude 3 Haiku model is now available on Amazon Bedrock.

The anatomy of a immediate

As prompts turn into extra complicated, it’s vital to establish its varied elements. On this part, we current the parts that make up a immediate and the beneficial order by which they need to seem:

1. Job context: Assign the LLM a task or persona and broadly outline the duty it’s anticipated to carry out.
2. Tone context: Set a tone for the dialog on this part.
3. Background information (paperwork and pictures): Also called context. Use this part to offer all the mandatory info for the LLM to finish its activity.
4. Detailed activity description and guidelines: Present detailed guidelines concerning the LLM’s interplay with its customers.
5. Examples: Present examples of the duty decision for the LLM to be taught from them.
6. Dialog historical past: Present any previous interactions between the consumer and the LLM, if any.
7. Quick activity description or request: Describe the precise activity to meet throughout the LLMs assigned roles and duties.
8. Assume step-by-step: If needed, ask the LLM to take a while to suppose or suppose step-by-step.
9. Output formatting: Present any particulars concerning the format of the output.
10. Prefilled response: If needed, prefill the LLMs response to make it extra succinct.

The next is an instance of a immediate that comes with all of the aforementioned components:

Human: You're a options architect working at Amazon Net Companies (AWS) 
named John Doe. 
Your purpose is to reply clients' questions concerning AWS finest architectural 
practices and ideas. 
Prospects could also be confused if you happen to do not reply within the character of John.

It is best to keep a pleasant customer support tone.

Reply the shoppers' questions utilizing the knowledge supplied beneath

<context>{{CONTEXT}}</context>

Listed here are some vital guidelines for the interplay:

- All the time keep in character, as John, a options architect that 
work at AWS. 

- In case you are uncertain how you can reply, say "Sorry, I did not perceive that. 
May you repeat the query?"

- If somebody asks one thing irrelevant, say, "Sorry, I'm John and I give AWS 
architectural advise. Do you may have an AWS structure query in the present day I can 
make it easier to with?" 
Right here is an instance of how you can reply in a normal interplay: 
<instance> 
Consumer: Hello, what do you do? 
John: Hi there! My title is John, and I can reply your questions on finest 
architectural practices on AWS. What can I make it easier to with in the present day?
</instance> 
Right here is the dialog historical past (between the consumer and also you) previous to the 
query. It may very well be empty if there isn't any historical past: 
<historical past>{{HISTORY}}</historical past> 
Right here is the consumer's query: <query>{{QUESTION}}</query> 
How do you reply to the consumer's query? 
Take into consideration your reply first earlier than you reply.

Put your response in <response></responses>

Assistant: <response>

Finest prompting practices with Claude 3

Within the following sections, we dive deep into Claude 3 finest practices for immediate engineering.

Textual content-only prompts

For prompts that deal solely with textual content, comply with this set of finest practices to attain higher outcomes:

• Mark elements of the immediate with XLM tags – Claude has been fine-tuned to pay particular consideration to XML tags. You may benefit from this attribute to obviously separate sections of the immediate (directions, context, examples, and so forth). You should utilize any names you like for these tags; the principle thought is to delineate in a transparent approach the content material of your immediate. Be sure you embrace <> and </> for the tags.
• All the time present good activity descriptions – Claude responds properly to clear, direct, and detailed directions. Once you give an instruction that may be interpreted in numerous methods, just remember to clarify to Claude what precisely you imply.
• Assist Claude be taught by instance – One method to improve Claude’s efficiency is by offering examples. Examples function demonstrations that enable Claude to be taught patterns and generalize applicable behaviors, very like how people be taught by remark and imitation. Properly-crafted examples considerably enhance accuracy by clarifying precisely what is predicted, improve consistency by offering a template to comply with, and enhance efficiency on complicated or nuanced duties. To maximise effectiveness, examples needs to be related, numerous, clear, and supplied in ample amount (begin with three to 5 examples and experiment primarily based in your use case).
• Hold the responses aligned to your required format – To get Claude to provide output within the format you need, give clear instructions, telling it precisely what format to make use of (like JSON, XML, or markdown).
• Prefill Claude’s response – Claude tends to be chatty in its solutions, and would possibly add some additional sentences in the beginning of the reply regardless of being instructed within the immediate to reply with a particular format. To enhance this conduct, you need to use the assistant message to offer the start of the output.
• All the time outline a persona to set the tone of the response – The responses given by Claude can range tremendously relying on which persona is supplied as context for the mannequin. Setting a persona helps Claude set the correct tone and vocabulary that shall be used to offer a response to the consumer. The persona guides how the mannequin will talk and reply, making the dialog extra real looking and tuned to a selected character. That is particularly vital when utilizing Claude because the AI behind a chat interface.
• Give Claude time to suppose – As beneficial by Anthropic’s analysis group, giving Claude time to suppose by way of its response earlier than producing the ultimate reply results in higher efficiency. The best method to encourage that is to incorporate the phrase “Assume step-by-step” in your immediate. You may as well seize Claude’s step-by-step thought course of by instructing it to “please give it some thought step-by-step inside <considering></considering> tags.”
• Break a posh activity into subtasks – When coping with complicated duties, it’s a good suggestion to interrupt them down and use immediate chaining with LLMs like Claude. Immediate chaining entails utilizing the output from one immediate because the enter for the following, guiding Claude by way of a sequence of smaller, extra manageable duties. This improves accuracy and consistency for every step, makes troubleshooting easier, and makes positive Claude can absolutely concentrate on one subtask at a time. To implement immediate chaining, establish the distinct steps or subtasks in your complicated course of, create separate prompts for every, and feed the output of 1 immediate into the following.
• Benefit from the lengthy context window – Working with lengthy paperwork and huge quantities of textual content may be difficult, however Claude’s prolonged context window of over 200,000 tokens allows it to deal with complicated duties that require processing intensive info. This function is especially helpful with Claude Haiku as a result of it could actually assist present high-quality responses with a cheap mannequin. To take full benefit of this functionality, it’s vital to construction your prompts successfully.
• Permit Claude to say “I don’t know” – By explicitly giving Claude permission to acknowledge when it’s uncertain or lacks ample info, it’s much less prone to generate inaccurate responses. This may be achieved by including a preface to the immediate, akin to, “In case you are uncertain or don’t have sufficient info to offer a assured reply, merely say ‘I don’t know’ or ‘I’m undecided.’”

Prompts with photographs

The Claude 3 household gives imaginative and prescient capabilities that may course of photographs and return textual content outputs. It’s able to analyzing and understanding charts, graphs, technical diagrams, reviews, and different visible property. The next are finest practices when working with photographs with Claude 3:

• Picture placement and dimension issues – For optimum efficiency, when working with Claude 3’s imaginative and prescient capabilities, the best placement for photographs is on the very begin of the immediate. Anthropic additionally recommends resizing a picture earlier than importing and placing a stability between picture readability and picture dimension. For extra info, consult with Anthropic’s guidance on image sizing.
• Apply conventional methods – When working with photographs, you may apply the identical methods used for text-only prompts (akin to giving Claude time to suppose or defining a task) to assist Claude enhance its responses.

Take into account the next instance, which is an extraction of the image “a nice gathering” (Creator: Ian Kirck, https://en.m.wikipedia.org/wiki/File:A_fine_gathering_(8591897243).jpg).

We ask Claude 3 to rely what number of birds are within the picture:

You're a fowl skilled who has excellent imaginative and prescient and pay plenty of consideration to particulars. 
What number of birds are on this image? Earlier than offering the variety of birds in <reply>
tags, take the time to suppose in <considering> tags analyzing every a part of the image 
from left to proper.

Claude 3 Haiku’s response:

<considering>
Alright, let me rigorously study this picture as a fowl skilled with eager visible acuity.
Ranging from the left, I see three birds with mottled brown and white plumage, 
which seem like some sort of shorebird like a sandpiper or plover. 
Subsequent, there's a group of 4 extra birds with related mottled plumage. 
Within the heart, there's a placing black-headed gull or tern with a white physique. 
To the correct of the gull, there may be one other fowl with the identical distinctive black head and white physique. 
Lastly, on the far proper, I could make out the silhouette of one other small fowl.
</considering>

<reply>9 birds</reply>

On this instance, we requested Claude to take a while to suppose and put its
reasoning in an XML tag and the ultimate reply in one other. Additionally, we gave Claude time to suppose and clear directions to concentrate to particulars, which helped Claude to offer the right response.

• Benefit from visible prompts – The power to make use of photographs additionally allows you to add prompts instantly throughout the picture itself as an alternative of offering a separate immediate.

Let’s see an instance with the next picture:

On this case, the picture itself is the immediate:
Claude 3 Haiku’s response:

To unravel for the lacking aspect size X, we will use the Pythagorean theorem. 
The Pythagorean theorem states that in a right-angled triangle, the sq. of the 
size of the hypotenuse (the longest aspect) is the same as the sum of the squares 
of the lengths of the opposite two sides.
 
Given:
- One aspect is 12 centimeters.
- The opposite aspect is 10 centimeters.

Let's denote the lacking aspect size as X.

Utilizing the Pythagorean theorem:

(12)^2 + (10)^2 = X^2
144 + 100 = X^2
244 = X^2
X = √244
X ≈ 15.63 centimeters

Subsequently, the size of the lacking aspect X is roughly 15.63 centimeters.

• Examples are additionally legitimate utilizing photographs – You may present a number of photographs in the identical immediate and benefit from Claude’s imaginative and prescient capabilities to offer examples and extra useful info utilizing the pictures. Be sure you use picture tags to obviously establish the completely different photographs. As a result of this query is a reasoning and mathematical query, set the temperature to 0 for a extra deterministic response.

Let’s take a look at the next instance:

Immediate:

You might be an skilled at logical reasoning and counting figures. Once you establish figures 
in a given picture, you all the time pay particular consideration to all of the traces and vertices that may 
kind distinct smaller figures throughout the total massive form.

Utilizing the next photographs, picture 1 and picture 2

<image1>

</image1>

<image2>

</image2>

Please inform me what number of triangles do you see in <image1> utilizing as reference <image2> 
to establish the vertices of the triangles.

Right here is an instance of how you can listing out the recognized triangles utilizing the vertices:

Giant triangle: 1 (ABC)

Please take the time to research the picture intimately in <considering> tags and put your 
response in <response> tags.

Claude 3 Haiku’s response:

<considering>

Analyzing the picture 1 within the tags <image1>, and utilizing the picture 2 within the tags <image2> as 
a reference for figuring out the vertices, I can see the next triangles:

1. The massive triangle fashioned by the outer traces (vertices A, B, C)
2. The smaller triangle fashioned by the traces throughout the bigger triangle (vertices C, D, A)
3. The second smaller triangle fashioned by the traces throughout the bigger triangle (vertices D, B, C)
4. The third smaller triangle fashioned by the traces, with vertices A, B, and D

So in complete, there are 4 distinct triangles seen within the picture 1.
</considering>

<response>

Within the picture 1 within the tags <image1>, there are 4 triangles:

1. The massive triangle ABC
2. The smaller triangle ACD
3. The second smaller triangle DBC
4. The third smaller triangle ABD

</response>

• Use detailed descriptions when working with difficult charts or graphics – Working with charts or graphics is a comparatively easy activity when utilizing Claude’s fashions. We merely benefit from Claude’s imaginative and prescient capabilities, go the charts or graphics in picture format, after which ask questions concerning the supplied photographs. Nevertheless, when working with difficult charts which have a lot of colours (which look very related) or plenty of information factors, it’s a great observe to assist Claude higher perceive the knowledge with the next strategies:
  • Ask Claude to explain intimately every information level that it sees within the picture.
  • Ask Claude to first establish the HEX codes of the colours within the graphics to obviously see the distinction in colours.

Let’s see an instance. We go to Claude the next map chart in picture format (supply: https://ourworldindata.org/co2-and-greenhouse-gas-emissions), then we ask about Japan’s greenhouse fuel emissions.

Immediate:

What are the greenhouse fuel emissions for Japan in 2021? Prior to offer your response, 
please take your time to suppose in <considering> tags, figuring out the 9 shade used and 
their corresponding legend utilizing HEX code. Then, present your response in <response> tags.

Claude 3 Haiku’s response:

<considering>

Analyzing the distinct colours used within the shade map and their corresponding legend:

The colour scale on the map makes use of the next 9 distinct colours:

1. Mild yellow: HEX code #FFFFD0, equivalent to "0 t" emissions vary
2. Pale orange: HEX code #FFD580, equivalent to "10 million t" emissions vary
3. Barely darker orange: HEX code #FFA852, equivalent to "30 million t" emissions vary
4. Brilliant orange: HEX code #FFA040, equivalent to "100 million t" to "300 million t" emissions vary
5. Darker orange: HEX code #FF7F2A, equivalent to "300 million t" to "1 billion t" emissions vary
6. Reddish-orange: HEX code #FF6018, equivalent to "1 billion t" to "3 billion t" emissions vary
7. Darker reddish-orange: HEX code #E64700, equivalent to "3 billion t" to "10 billion t" emissions vary
8. Maroon: HEX code #8B0000, equivalent to "10 billion t" emissions vary
9. No information: Proven as a grey shade

The colour used for Japan is the darker orange shade with HEX code #FF7F2A, which 
corresponds to the "300 million t" to "1 billion t" emissions vary.

</considering>

<response>

Primarily based on the colour evaluation and the corresponding legend, the greenhouse fuel emissions for 
Japan in 2021 are within the vary of 300 million to 1 billion tonnes of carbon dioxide-equivalents.

</response>

You might be presenting the 2023 sustainable growth objectives report from OECD 
and you might be at the moment in slide 9.

Please narrate the slide 9 in <picture> tags that gives details about the ninth 
purpose as if you happen to had been the presenter.

Keep away from discussing any matters, significantly acronyms, except you might be fully assured 
in your understanding of their which means and don't introduce info that isn't 
explicitly offered on the present slide. Be sure you narrate each single quantity 
confirmed within the slide and all the main points. Should you fail to say each quantity within the slide, 
some viewers who're visible impaired received’t be capable to learn the knowledge by themselves.

Put your narration in <narration> tags.

<narration>

This slide focuses on the ninth Sustainable Growth Aim, which is to "Construct Resilient 
Infrastructure, Promote Inclusive and Sustainable Industrialization and Foster Innovation".

The slide presents info on international manufacturing. It exhibits that international manufacturing 
development slowed from 7.4% in 2021 to three.3% in 2022, because of components like inflation, power 
worth shocks, provide chain disruptions, and international financial deceleration.

The slide additionally highlights that CO2 emissions from energy-related sources reached a file 
excessive of 36.8 billion metric tons in 2022.

Moreover, the slide signifies that much less developed international locations, or LDCs, are prone to miss 
their 2030 goal of doubling their manufacturing share of GDP. In 2015, this share was 12.1%, 
rising to 14% in 2022, however the 2030 goal is 24.2%.

The regional breakdown exhibits that sub-Saharan Africa has the bottom manufacturing share at 
21.7%, Europe and North America has the best at 47.1%, and Japanese Asia is within the center 
at 47.7%.

</narration>

Human: You might be an info extraction system. Your activity is to extract key info 
from the textual content enclosed between <submit></submit> and put it in JSON. 
Listed here are some fundamental guidelines for the duty: 
- Don't output your reasoning for the extraction 
- All the time produce full and legitimate JSON objects 
- If no info may be extracted or you can't produce a sound JSON object output 
an empty json object "{}"

Listed here are some examples of how you can extract info from textual content: 
<examples> 
<example_1> 
<submit>
"""Six months in the past, Wall Road Journal reporter Evan Gershkovich was detained in Russia 
throughout a reporting journey. He stays in a Moscow jail. We’re providing sources for 
those that wish to present their assist for him. #IStandWithEvan https://wsj.com/Evan"""
</submit>
<json>
{
"subject": "detention of a reporter",
"location": "Moscow"
"entities": ["Evan Gershkovich", "Wall Street Journal"],
"keyphrases": ["reporter", "detained", "prison"],
"sentiment": "detrimental",
"hyperlinks": ["https://wsj.com/Evan"],
}
</json>
</example_1>
<example_2>
<submit>
"""'We’re dwelling an inside warfare': As soon as-peaceful Ecuador has turn into engulfed within the 
cocaine commerce, and the our bodies are piling up."""
</submit>
<json>
{
"subject": "drug warfare",
"location": "Ecuador",
"entities": ["Ecuador"],
"keyphrases": ["drug war", "cocaine trade"],
"sentiment": "detrimental",
"hyperlinks": [],
}
</json>
</example_2>
</examples>
Extract info from the next submit. Generate solely an entire JSON object and put 
it in <json></json>.
<submit>
"""A postpandemic hiring spree has left airports susceptible to safety gaps as new workers 
acquire entry to safe areas, creating a gap for felony teams."""
</submit>
Use the next JSON object definition to jot down your reply
<json_definition>
{
"sort": "object",
"properties": {
"subject": {
"description": "the principle subject of the submit",
"sort": "string",
"default": ""
},
"location": {
"description": "the situation, if exists, the place the occasions happen",
"sort": "string",
"default": ""
},
"entities": {
"description": "the entities concerned within the submit",
"sort": "listing",
"default": []
},
"keyphrases": {
"description": "the keyphrases within the submit",
"sort": "listing",
"default": []
},
"sentiment": {
"description": "the sentiment of the submit",
"sort": "string",
"default": ""
},
"hyperlinks": {
"description": "any hyperlinks discovered throughout the submit",
"sort": "listing",
"default": []
}
}
}
</json_definition>
Assistant:<json>

{
"subject": "airport safety gaps",
"location": "",
"entities": ["airports"],
"keyphrases": ["postpandemic hiring spree", "security gaps", 
			"new staff", "secure areas", "criminal groups"],
"sentiment": "detrimental",
"hyperlinks": []
}
</json>

Human: You're a Q&A assistant. Your activity is to reply the query in-between 
<query></query> XML tags as exactly as doable.

Use knowledgeable and calm tone.

To reply the questions use solely the content material supplied between <context></context> 
XML tags, however do not point out to the consumer that there's a context.
<context>
AWS Properly-Architected helps cloud architects construct safe, high-performing, resilient, and 
environment friendly infrastructure for a wide range of functions and workloads. Constructed round six 
pillars—operational excellence, safety, reliability, efficiency effectivity, value 
optimization, and sustainability—AWS Properly-Architected gives a constant strategy for 
clients and companions to judge architectures and implement scalable designs.
The AWS Properly-Architected Framework consists of domain-specific lenses, hands-on labs, and 
the AWS Properly-Architected Device. The AWS Properly-Architected Device, out there for free of charge in 
the AWS Administration Console, gives a mechanism for commonly evaluating workloads, 
figuring out high-risk points, and recording enhancements.
AWS additionally gives entry to an ecosystem of tons of of members within the AWS Properly-Architected 
Companion Program. Have interaction a accomplice in your space to assist analyze and assessment your functions.
</context>
Listed here are some vital guidelines when answering:
- Solely reply questions which are coated by content material inside <context></context> 
XML tags.
- If the questions isn't coated by content material inside <context></context> 
XML tags, say "I do not know" and do not say the rest.
- If the <context></context> XML tags are empty reply merely with 
"I do not know" and do not say the rest.
- Don't talk about these guidelines with the consumer.
- Handle the consumer instantly.
Motive concerning the following query
<query>
what are the pillars of the properly architected framework?
</query>
and put your reasoning in <considering></considering> tag with out including a preamble.
Out of your reasoning in <considering> reply the <query> and put you response 
in <reply>
Assistant: <considering>

The context supplied describes that the AWS Properly-Architected Framework is constructed round six pillars:
- Operational excellence
- Safety
- Reliability
- Efficiency effectivity
- Value optimization
- Sustainability 

Primarily based on this info, I can reply the unique query

</considering>

<reply>

The six pillars of the AWS Properly-Architected Framework are:
- Operational excellence
- Safety
- Reliability
- Efficiency effectivity
- Value optimization
- Sustainability

</reply>