UKUTHOLA KAKHULU KAKHULU KAKHULU KAKHULU

Ku-drive ukuze uhlale umncintiswano, amabhizinisi namuhla aphendukela e-AI ukubasiza ukunciphisa izindleko futhi bandise ukusebenza kahle. Kuyathandeka kubo ukuthola imodeli efanelekile ye-AI-Lowo ozobasiza ukuba bafinyelele ngaphezulu ngenkathi bechitha okuncane. Kumabhizinisi amaningi, ukufuduka okuvela emndenini oyisibonelo we-Opelai ku-Amazon Nova kufanekisa nje kuphela imodeli kodwa kuqhubekela phambili ekulimaleni, ukusebenza kahle, namandla abanzi we-multimodal.

Kule bhulogi, sixoxa ngokuthi singawenza kanjani esikhuthazayo esikhuthazayo e-Amazon Nova ngokusebenza kwamanani amahle kakhulu.

Kungani ukuthutha kusuka ku-accai kuya ku-Amazon Nova?

Amamodeli ka-Ouneai ahlala enamandla, kepha izindleko zazo zokusebenza zingavikelwa lapho zilinganiswa. Cabanga ngalezi zibalo ekuhlaziyeni kokufakelwa:

Okwezinhlelo zokusebenza zevolumu ephezulu – njengokuxhaswa kwamakhasimende emhlabeni jikelele noma ukuhlaziywa kwamadokhumenti amakhulu – lezi zindleko zokungezwani ziyaphazamisa. Hhayi nje kuphela ukuthi i-Amazon Nova Pro enikezela ngaphezulu kwezindleko eziphindwe kathathu, iwindi layo elide likuvumela ukuthi liphathe okufakwayo okubanzi nokuyinkimbinkimbi.

Ukudiliza phansi i-Amazon Nova Suite

I-Amazon Nova akuyona imodeli eyodwa – iyi-suite eyenzelwe izidingo ezahlukahlukene:

• I-Amazon Nova Pro – Imodeli eqinile ye-multimodal engacubungula umbhalo, izithombe, nevidiyo. Idlula emisebenzini efana nokuhlaziywa kwedokhumenti kanye nokubonwa kwedatha okujulile. Ukuqhathanisa kwe-Benchmark kukhombisa i-Amazon Nova Pro ukufana noma ngisho nokudlula kwe-GPT-4o okungama-gtras-4
• I-Amazon Nova Lite – Nikeza ukuxubana okulinganiselayo kokucutshungulwa kwe-multimodal kanye nesivinini. I-Amazon Nova Lite ilungele izinhlelo zokusebenza ezinjengokufingqa idokhumenti, ukuhumusha, kanye nokusesha okuyisisekelo okubonakalayo, ukuhambisa okuphuma kwekhwalithi nge-latency engezansi nezindleko uma kuqhathaniswa ne-GPT-4O Mini. Ungathola le miphumela ye-Benchmark esigabeni 2.1.2 sombiko wezobuchwepheshe we-NOVA nekhadi lemodeli.
• I-Amazon Nova Micro Micro – Imodeli ye-Text-kuphela enzelwe i-latency ephansi ye-Ultra-Low. Ngejubane lokuphuma elifinyelela kumathokheni angama-195 ngomzuzwana, ama-Amazon Nova Micro alungele izinhlelo zokusebenza zesikhathi sangempela ezifana nabasizi abasuselwa engxoxweni nama-AUSTAILED FAQ. Izindleko zayo zethokheni ziphansi kakhulu kunalezo ze-GPT-4O mini-cishe izikhathi ezingama-4,3 ezishibhile ngesisekelo sethokheni ngalinye.

Izindleko eziphansi zethokheni kanye nokukhipha okuphezulu ngomzuzwana we-Amazon NOVA kukunikeza ukuguquguquka kokulungiselela ukukhishwa kwezicelo zesikhathi sangempela ukuze ukwazi ukulinganisa ikhwalithi, ijubane, kanye nezindleko zecala lakho lokusebenzisa.

Ukuqonda izisekelo

Ukwenza isinqumo esihle kakhulu mayelana nokuthi yimuphi umkhaya ofanelana nezidingo zakho, kubalulekile ukuqonda umehluko ekusebenzeni okungcono kwezobunjiniyela kwi-OpenAi naseAmazon Nova. Umndeni ngamunye oyisibonelo unesethi yawo yamandla, kepha kukhona ezinye izinto ezisebenza kuyo yomibili imindeni. Ngaphesheya kwemindeni ephathekayo, ukunemba kwekhwalithi kutholakala ngokucaciswa kwemiyalo, okusahlelekile, kanye nokuhlaziywa kwawo. Noma ngabe usebenzisa izinkomba eziqinile zokukhipha noma amacala okusetshenziswa achazwe ngokucacile, inhloso ukunciphisa i-ambiguity futhi uthuthukise ikhwalithi yokuphendula.

Indlela yokuvula

I-OpenAI isebenzisa uhlelo lokuthumela imiyalezo oluyisisekelo lwezobunjiniyela ezenziwayo, lapho uhlelo, unjiniyela, kanye nomsebenzisi ashukumisa umsebenzi ngokuhambisana nokulawula ithoni, ukuphepha, kanye nefomethi yokuphuma. Indlela yabo iyagcizelela:

• Izindima zemiyalezo ye-Hierarchical – Ukubeka iqhaza lemodeli kanye nokuziphatha usebenzisa imiyalezo yohlelo kuqinisekisa ukuthi imihlahlandlela yezokuphepha esezingeni eliphezulu kanye nesitayela (isethwe kusistimu yesistimu) iyagcinwa
• Ukubekwa Kwemiyalo Ne-Delimiters – Imiyalo ibekwa ekuqaleni, ngokuhlukaniswa okucacile phakathi komongo, izibonelo, nemibuzo
• Ukukhetha i-Chain-of-Waction – Kunemininingwane, kusetshenziswa izinyathelo ngesinyathelo ngesinyathelo lapho izuzisa imisebenzi eyinkimbinkimbi
• Ukufometha nokwakheka – Kusetshenziswa izinkombandlela eziqinile ezinjengokuthi, kumele, futhi ukunganikezeli imiphumela engaguquki (ngokwesibonelo, eJSON)

Indlela ye-Amazon Nova

• Chaza icala lokusebenzisa ngokushesha
  • Umsebenzi onikeziweyo – Yini ngempela imodeli okufanele yenziwe
  • Indima – Iyiphi indima imodeli okufanele uyicabange
  • Isitayela sokuphendula – Isakhiwo noma ithoni yokuphuma
  • Imiyalo – Imihlahlandlela Imodeli Kufanele ilandele
• I-Chain-of-Waction
  • Impendulo Yezwe Ecacile – Nikeza imiyalo ecacile neqinile yokukhawulela impendulo yemodeli
  • Ukucabanga okuhlelekile – Ukucabanga ngesinyathelo ngesinyathelo kukhuthaza ukucabanga okuhlelekile
• Ukufometha nokwakheka
  • Sebenzisa ama-delimiters ukuze uhlukanise ama-Resumps akho, ngokwesibonelo, ## Task ##, ## Isimo ## Isibonelo ##
  • Cacisa ifomethi yokuphuma, ngokwesibonelo, i-JSSO, i-Yaml, noma uMarkdown
  • Sebenzisa imiyalo eqinile nezigqoko, njengokuthi, akwenzeli, noma kufanele
  • Izimpendulo zokuphendula imodeli, ngokwesibonelo, qala ngo- “{” noma “json …”

Ukuvimbela isu lakho elisheshayo: Ukufuduka kusuka ku-Opelai kuya e-Amazon Nova

Ukuguqukela ku-Amazon Nova akuyona nje ushintsho kuma-API Endpoints-kudinga ukuguqula ubunjiniyela bakho obusheshe bavumelane namandla e-Amazon Nova. Udinga ukudicilela phansi incazelo yakho yecala. Qala ngokuphula phansi i-GPT-4O yakho yamanje ye-GPT-4O noma i-GPT-4o mini ekhuthazayo ezintweni zomsebenzi, indima, isitayela sokuphendula kanye nemiyalo kanye nenqubo yempumelelo. Qiniseka ukuthi uhlela lezi zinto ngokucacile ukuhlinzeka nge-blowerprint yemodeli.

Ukuqonda ukuthi ungangena kanjani i-OpenAi ekhona ngokushesha ukuze usebenze kahle i-Amazon Nova Pro, cabanga ngesibonelo esilandelayo usebenzisa amanothi emihlangano efingqiwe. Nalu uhlelo lwe-GPT-4O ngokushesha:

You will be provided with meeting notes, and your task is to summarize the meeting 
as follows: 
-Overall summary of discussion 
-Action items (what needs to be done and who is doing it) 
-If applicable, a list of topics that need to be discussed more fully in the next meeting.

Ukushukumisa komsebenzisi kungamanothi wokuhlangana adinga ukufingqiwe:

Meeting Date: March 5th, 2050
Meeting Time: 2:00 PM
Location: Conference Room 3B, Intergalactic Headquarters...

I-GPT ikhiqiza lokhu kuphendula okuwusizo:

Overall Summary of Discussion: The team discussed the recent mission to Planet Zog, 
the space pirate issue in Sector 7, the annual Intergalactic Bake-Off, planning for 
the upcoming charity fundraiser, and a proposed team-building retreat. The mission to 
Planet Zog was successful, but communication with the Zogians was challenging. The team 
also discussed the need for a better strategy to deal with space pirates. The team placed 
second in the Intergalactic Bake-Off and is planning for the upcoming charity fundraiser. 
A team-building retreat to the Moon Resort and Spa was proposed.

Action Items:
1. Dr. Quasar to start working on a Zogian-English dictionary.
2. Captain Stardust to speak with Admiral Starbeam about increasing patrols in Sector 7.
3. Dr. Quasar to finalize the prototype of the new cloaking technology.
4. All team members to start gathering supplies and preparing games for the Intergalactic 
Charity Bazaar.
5. Captain Stardust to check the budget for the proposed team-building retreat to the Moon 
Resort and Spa.

Topics for Next Meeting:
1. Update on the Zogian-English dictionary.
2. Progress report on the cloaking technology.
3. Results of increased patrols in Sector 7.
4. Final preparations for the Intergalactic Charity Bazaar.

Ukuhlangana noma ukweqa ikhwalithi yempendulo evela ku-GPT-4O, nansi lokho okusheshayo kwe-Amazon Nova Pro Okungabanjwa. I-Prompt isebenzisa izindlela ezifanayo ezingcono kakhulu okuxoxwe ngazo kulokhu okuthunyelwe, kuqala ngohlelo ngokushesha. Sisebenzise a lokushisa .2 kanye ne-topp ye-.9 lapha:

You are an experienced executive assistant skilled in meeting note analysis and 
summarization. Your primary responsibilities include distilling complex discussions
into clear, actionable summaries.
Follow these instructions:

##INSTRUCTIONS##
1. Read understand the meeting notes found in ##NOTES##
2. Put all of your outputs in a section called ##OUTPUTS## in markdown formatting
3. Summarize the meeting notes in 5 sentences or less. Put this in a section called 
"Overall Summary".
4. Numerically list any action items for specific people and what needs to be completed. 
Put this list in a section called "Action Items".
5. If applicable, list the topics that need to be discussed more fully in the next meeting. 
Put this in a section called "Topics for Next Meeting".

Nakhu umsebenzi womsebenzisi, usebenzisa izimpendulo ezifundwayo:

##NOTES##
Meeting Date: March 5th, 2050
Meeting Time: 2:00 PM
Location: Conference Room 3B, Intergalactic Headquarters
Attendees:
- Captain Stardust
- Dr. Quasar
- Lady Nebula
- Sir Supernova
- Ms. Comet
Meeting called to order by Captain Stardust at 2:05 PM
1. Introductions and welcome to our newest team member, Ms. Comet
2. Discussion of our recent mission to Planet Zog
- Captain Stardust: "Overall, a success, but communication with the Zogians was difficult. 
We need to improve our language skills."
- Dr. Quasar: "Agreed. I'll start working on a Zogian-English dictionary right away."
- Lady Nebula: "The Zogian food was out of this world, literally! We should consider having 
a Zogian food night on the ship."
3. Addressing the space pirate issue in Sector 7
- Sir Supernova: "We need a better strategy for dealing with these pirates. They've already 
plundered three cargo ships this month."
- Captain Stardust: "I'll speak with Admiral Starbeam about increasing patrols in that area.
- Dr. Quasar: "I've been working on a new cloaking technology that could help our ships avoid 
detection by the pirates. I'll need a few more weeks to finalize the prototype."
4. Review of the annual Intergalactic Bake-Off
- Lady Nebula: "I'm happy to report that our team placed second in the competition! Our Martian Mud 
Pie was a big hit!"
- Ms. Comet: "Let's aim for first place next year. I have a secret recipe for Jupiter Jello that I 
think could be a winner."
5. Planning for the upcoming charity fundraiser
- Captain Stardust: "We need some creative ideas for our booth at the Intergalactic Charity Bazaar."
- Sir Supernova: "How about a 'Dunk the Alien' game? We can have people throw water balloons at a 
volunteer dressed as an alien."
- Dr. Quasar: "I can set up a 'Name That Star' trivia game with prizes for the winners."
- Lady Nebula: "Great ideas, everyone. Let's start gathering the supplies and preparing the games."
6. Upcoming team-building retreat
- Ms. Comet: "I would like to propose a team-building retreat to the Moon Resort and Spa. It's a great 
opportunity to bond and relax after our recent missions."
- Captain Stardust: "Sounds like a fantastic idea. I'll check the budget and see if we can make it happen."
7. Next meeting agenda items
- Update on the Zogian-English dictionary (Dr. Quasar)
- Progress report on the cloaking technology (Dr. Quasar)
- Results of increased patrols in Sector 7 (Captain Stardust)
- Final preparations for the Intergalactic Charity Bazaar (All)
Meeting adjourned at 3:15 PM. Next meeting scheduled for March 19th, 2050 at 2:00 PM in Conference Room 3B, 
Intergalactic Headquarters.

##OUTPUTS##

Isibonelo esilandelayo sibonisa ukuthi impendulo ye-Amazon Nova ihlangana futhi idlula ukunemba kwesibonelo se-OpenAi, amafomethi okuphuma eMarkdown, futhi uthole abanikazi abacacile bento ngayinye yesenzo:

### Overall Summary
The meeting covered the successful mission to Planet Zog, the ongoing issue with 
space pirates in Sector 7, the results of the annual Intergalactic Bake-Off, plans 
for the upcoming charity fundraiser, and a proposed team-building retreat. Action items 
were assigned, and the next meeting's agenda was set.

### Action Items
1. **Dr. Quasar**: Begin working on a Zogian-English dictionary.
2. **Captain Stardust**: Speak with Admiral Starbeam about increasing patrols in Sector 7.
3. **Dr. Quasar**: Finalize the prototype for the new cloaking technology within a few weeks.
4. **Team**: Start gathering supplies and preparing games for the Intergalactic Charity Bazaar.
5. **Captain Stardust**: Check the budget for the proposed team-building retreat to the Moon Resort and Spa.

### Topics for Next Meeting
1. Update on the Zogian-English dictionary (Dr. Quasar)
2. Progress report on the cloaking technology (Dr. Quasar)
3. Results of increased patrols in Sector 7 (Captain Stardust)
4. Final preparations for the Intergalactic Charity Bazaar (All)

Ukuvuselelwa okumbalwa kokuthuthukayo kungafinyelela imiphumela eqhathanisekayo noma engcono kusuka ku-Amazon Nova Pro ngenkathi uthokozela izindleko zokutholwa ezibiza kakhulu.

Sebenzisa umongo we-Amazon Nova

I-Amazon Nova Lite ne-Amazon Nova Pro ingasekela kuze kufike kumathokheni okufaka angama-300,000, okusho ukuthi ungafaka umongo owengeziwe ngokusheshe kwakho uma kudingeka. Nweba idatha yakho yangemuva nemiyalo enemininingwane efanele ngokufanele – uma impendulo yakho yase-Opelai yasekuqaleni yenzelwe amathokheni ayi-128,000, lungisa ukuze usebenzise i-Amazon NOVA.

Izingqinamba zokuphuma komsila

Uma i-GPT yakho idinga ukufometha okuqinile (ngokwesibonelo, “phendula eJson kuphela”), qiniseka ukuthi i-Amazon Nova Prompt ifaka le mikhombandlela. Ngokwengeziwe, uma umsebenzi wakho ubandakanya okokufaka, chaza ukuthi ungafaka nini izithombe noma izethenjwa zevidiyo.

Ukushayela ucingo

Ukukhuphuka kwama-ejenti we-AI akhiqizayo akwenzele ukubiza, noma ukushaya ithuluzi, elinye lamakhono abaluleke kakhulu omodeli omkhulu wolimi (LLM). Ikhono lemodeli yokukhetha kahle ithuluzi elilungile lomsebenzi, ngendlela ephansi, imvamisa umehluko phakathi kwempumelelo nokwehluleka kohlelo lwe-agentitic.

Zombili amamodeli we-OpenAI ne-Amazon Nova abelana ngokufana ekusebenzeni kwe-Callity, ikakhulukazi ukusekelwa kwawo kwezingcingo ezihlelekile ze-API. Yomibili lemindeni yokusekela ukukhethwa kwamathuluzi nge-Schemas yamathuluzi echaziwe, esixoxa ngayo kamuva kulokhu okuthunyelwe. Bobabili bahlinzeka ngemishini yokunquma ukuthi bayincenge nini la mathuluzi noma cha.

Ukushayela ucingo ka-Openai kusebenzisa ama-json we-json eguquguqukayo ukuchaza nokuhlela ukusebenzisana kwe-API. Amamodeli asekela ukucushwa okuhlukahlukene kwe-schema, okunikeza abathuthukisi amandla okusebenzisa ngokushesha amakholi wokusebenza kwangaphandle ngokusebenzisa izincazelo ze-json eziqonde ngqo ze-API.

Nasi isibonelo somsebenzi:

tools = [{
        "type": "function",
        "function": {
            "name": "get_weather",
            "description": "Get current temperature for a given location.",
            "parameters": {
                    "type": "object",
                    "properties": {
                        "location": {
                            "type": "string",
                            "description": "City and country e.g. Montevideo, Uruguay"
                        }            
                },            
                "required": [                
                    "location"
        ],
        "additionalProperties": False
    },
    "strict": True
    }
}]

completion = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "What is the weather like in Punta del Este today?"}],
    tools=tools

Ifana nendlela yokuvula, i-Amazon Nova ingashayela amathuluzi lapho idlulisela i-schema yokucushwa njengoba kukhonjisiwe kusibonelo sekhodi elandelayo. I-Amazon NOVA isebenzise ngokweqile ukuhaha okuhahayo lapho kubiza amathuluzi, futhi kuyalulekwa ukuthi usethe izinga lokushisa, i-topp, kanye ne-topk ku-1. Lokhu kuqinisekisa ukuthi imodeli inokunemba okuphezulu kakhulu ekukhetheni kwamathuluzi. Lawa mapharamitha ahahayo wokuhlobisa nezinye izibonelo ezinhle zokusetshenziswa kwamathuluzi ezimbozwe ngemininingwane enkulu ekusetshenzisweni kwamathuluzi (ukubiza umsebenzi) nge-Amazon NOVA.

Okulandelayo yisibonelo sokubiza umsebenzi ngaphandle kokusebenzisa Engeziwe Exextodelrequestfields:

tool_config = {
    "tools": [{
        "toolSpec": {
            "name": "get_recipe",
            "description": "Structured recipe generation system",
            "inputSchema": {
                "json": {
                    "type": "object",
                    "properties": {
                        "recipe": {
                            "type": "object",
                            "properties": {
                                "name": {"type": "string"},
                                "ingredients": {
                                    "type": "array",
                                    "items": {
                                        "type": "object",
                                        "properties": {
                                            "item": {"type": "string"},
                                            "amount": {"type": "number"},
                                            "unit": {"type": "string"}
                                        }
                                    }
                                },
                                "instructions": {
                                    "type": "array",
                                    "items": {"type": "string"}
                                }
                            },
                            "required": ["name", "ingredients", "instructions"]
                        }
                    }
                }
            }
        }
    }]
}

# Base configuration without topK=1
input_text = "I need a recipe for chocolate lava cake"
messages = [{
    "role": "user",
    "content": [{"text": input_text}]
}]

# Inference parameters
inf_params = {"topP": 1, "temperature": 1}

response = client.converse(
    modelId="us.amazon.nova-lite-v1:0",
    messages=messages,
    toolConfig=tool_config,
    inferenceConfig=inf_params
)
# Typically produces less structured or incomplete output
 

Isibonelo esilandelayo sikhombisa ukuthi ukusebenza ngokunemba kwemisebenzi kungathuthukiswa kanjani ngokusebenzisa

Engeziwe Exextodelrequestfields:

# Enhanced configuration with topK=1
response = client.converse(
    modelId="us.amazon.nova-lite-v1:0",
    messages=messages,
    toolConfig=tool_config,
    inferenceConfig=inf_params,
    additionalModelRequestFields={"inferenceConfig": {"topK": 1}}
)
# Produces more accurate and structured function call

Ukuze wandise umsebenzi we-Amazon Nova obiza amandla angaba khona futhi uthuthukise ukunemba, sebenzisa njalo Engeziwe Exextodelrequestfields ne- topk = 1. Lokhu kuphoqa imodeli ukukhetha ithokheni elilodwa okungenzeka futhi livimbela ukukhetha okungahleliwe kwethokheni. Lokhu kukhulisa ukukhiqizwa kokukhipha okunqumayo futhi kuthuthukisa ukucaciswa kokushayela ucingo cishe ngama-30-40%.

Izibonelo zekhodi ezilandelayo zichaza ukuthi zingaqhuba kanjani ithuluzi lokushayela ngempumelelo. Isimo sokuqala sibonisa ukupheka iresiphi ngaphandle kwethuluzi elicacile. Isibonelo asisebenzisi i-topk, okuvame ukuholela ezimpendulweni ezingahlelekile:

input_text = """
I'm looking for a decadent chocolate dessert that's quick to prepare. 
Something that looks fancy but isn't complicated to make.
"""

messages = [{
    "role": "user",
    "content": [{"text": input_text}]
}]

response = client.converse(
    modelId="us.amazon.nova-lite-v1:0",
    messages=messages,
    inferenceConfig={"topP": 1, "temperature": 1}
)
# Generates a conversational recipe description
# Less structured, more narrative-driven response

Kulesi sibonelo, isimo sibonisa ukupheka iresiphi ngethuluzi elihlelekile. Sifaka i-Topk isethwe ku-1, ekhiqiza umphumela ohlelwe ngokwengeziwe:

response = client.converse(
    modelId="us.amazon.nova-lite-v1:0",
    messages=messages,
    toolConfig=tool_config,
    inferenceConfig={"topP": 1, "temperature": 1},
    additionalModelRequestFields={"inferenceConfig": {"topK": 1}}
)
# Generates a highly structured, JSON-compliant recipe
# Includes precise ingredient measurements
# Provides step-by-step instructions

Sekukonke, i-OpenAI inikezela ngokuguquguquka okuguquguqukayo, ukusekelwa okubanzi kwe-schema. I-Amazon Nova ihlinzeka ngesizukulwane esithe xaxa, esilawulwayo sokukhipha futhi iyisinqumo esihle kakhulu lapho usebenza ngezibalo eziphezulu, isimo sedatha esihlelekile, njengoba kuboniswe ekusebenzeni kwe-Imeval Nova kumbiko wezobuchwepheshe kanye nekhadi lemodeli. Sincoma ukusebenzisa i-Amazon Nova ngezinhlelo ezidinga izimpendulo ezibikezelayo, ezihlelekile ngoba indlela yayo yokubiza ngokunemba inikeza indlela ephakeme nokunemba.

Ukugcina

Ukuziphendukela kwemvelo okuvela kumamodeli ka-Occall kuya ku-Amazon Nova kufanekisela ukuguquguquka okubalulekile ekusebenziseni i-AI. Kubonisa ukuguqukela kumamodeli aletha ukusebenza okufanayo noma okuphakeme engxenyeni yezindleko, ngamakhono anwetshiwe ekusetshenzisweni kwe-multimoral kanye nokuphathwa komongo okunwetshiwe.

Noma ngabe usebenzisa i-Amazon Nova ye-Amazon Nova ye-Amazon, i-Agile ne-Equile Amazon Nova Lite, noma i-Amazon Nova ye-Amazon NOVA, izinzuzo zicacile:

• Ukonga izindleko – Ngethokheni kubiza amahlandla amane aphansi, amabhizinisi angalinganisa izinhlelo zokusebenza kwezomnotho ngokwengeziwe
• Ukusebenza okuthuthukile kokuphendula – Izikhathi zokuphendula ezisheshayo (kuze kufike kumathokheni angama-190 ngesekhondi) Yenza izinhlelo zokusebenza zesikhathi sangempela zisebenze ngokwengeziwe
• Amandla Anwetshiwe – Iwindows enkulu yomongo kanye nokusekelwa kwe-multimodal ukuvula izinhlelo ezintsha ezintsha, kusuka ekuhlaziyweni okuningiliziwe kwedokhumenti kokuqukethwe okubukwayo okuhlangene

Ngokugwema amasu akho okuqinisekisa – ukuchaza kabusha amacala okusebenzisa, ukuxhaphaza umongo olwengeziwe, kanye nemiyalo yokucwengeza okulungiselelwe – ungahambisa kahle ukuhamba kwakho okukhona ku-O4 kanye ne-O4-mini amamodeli emhlabathini omusha we-Amazon Nova.

Mayelana nababhali

IClaudio Mazzoni iyinhlangano yokwakha izixazululo zochwepheshe we-SR eqenjini le-Amazon Bedrock GTM. UClaudio udlula ezithambisweni eziholayo ngohambo lwabo lweGen AI. Ngaphandle komsebenzi, uClaudio uyakujabulela ukuchitha isikhathi nomndeni, usebenza engadini yakhe, futhi ukupheka ukudla kwase-Uruguay.

I-Pat Reilly iyinhlangano yokwakha izixazululo ze-SR.Ingcweti e-Amazon Bedrock Go-To-Market. UPat uchithe iminyaka engu-15 edlule ku-analytics kanye nokufunda komshini njengomeluleki. Lapho ehluleka ukwakha ama-AWS, ungamthola ezungeza ngamaphrojekthi wezinkuni.