Gutegura Agentic AI mumiyoboro ya DevOps hamwe na DSPy na LangChain kuri ZeroOps Yongerewe

Intangiriro

Kwishyira hamwe kwa Artific Intelligence (AI) mu miyoboro ya DevOps byahinduye uburyo software ikora, igeragezwa, kandi ikoherezwa. Agentic AI, igice cya AI gifasha gufata ibyemezo byigenga no gufata ibyemezo, bifite ubushobozi bwo kurushaho kunoza imikorere nibikorwa bya DevOps. Muri iki kiganiro, tuzasesengura igitekerezo cyo gutunganya AI agent mu miyoboro ya DevOps dukoresheje DSPy na LangChain, tunaganira ku nyungu zubu buryo mu kugera kuri ZeroOps.

Agentic AI ni iki?

Agentic AI bivuga ubwoko bwa AI bufite ubushobozi bwo gufata ibyemezo no gufata ibyemezo byigenga, nta muntu ubigizemo uruhare. Ibi bigerwaho hifashishijwe imashini igezweho yiga algorithms, gutunganya ururimi karemano, hamwe nubwubatsi bwubwenge. Sisitemu ya Agentic AI irashobora kwigira kuburambe, guhuza nibihe bishya, kandi igasabana nabantu muburyo busanzwe kandi bwimbitse.

ZeroOps ni iki?

ZeroOps ni filozofiya ya DevOps igamije gukuraho ibikenewe kugira uruhare rwabantu mu iterambere rya software, kubohereza, no kubungabunga. Ibi bigerwaho hifashishijwe gukoresha automatike, orchestre, hamwe no gufata ibyemezo bya AI. ZeroOps ifasha amashyirahamwe gukoresha software byihuse, hamwe nubwiza buhanitse, kandi ku giciro gito.

Gutegura Agentic AI mu miyoboro ya DevOps

Kugirango dutegure AI ya agent mu miyoboro ya DevOps, dukeneye guhuza AI iterwa no gufata ibyemezo no gufata ingamba mubikorwa. Ibi birashobora kugerwaho hifashishijwe DSPy na LangChain, uburyo bubiri buzwi bwo kubaka no gukoresha porogaramu ikoreshwa na AI.

DSPy

DSPy ni Python urwego rwo kubaka no gukoresha porogaramu ikoreshwa na AI. Itanga urutonde rwibikoresho namasomero yo gutunganya amakuru, kwiga imashini, no gutunganya ururimi karemano. DSPy yagenewe gukora nta nkomyi hamwe nurwego ruzwi nka LangChain na FastAPI.

LangChain

LangChain ni Python urwego rwo kubaka no gukoresha porogaramu ikoreshwa na AI. Itanga urutonde rwibikoresho nububiko bwibitabo byo gutunganya ururimi karemano, kwiga imashini, hamwe nubwubatsi bwubwenge. LangChain yagenewe gukora ntakabuza hamwe nibikorwa bizwi nka DSPy na FastAPI.

Gushyira mu bikorwa tekinike

Kugaragaza ishyirwa mubikorwa rya tekiniki ya orchestre agentic AI mumiyoboro ya DevOps ukoresheje DSPy na LangChain, reka dusuzume urugero rworoshye. Dufate ko dushaka gukoresha itangwa rya porogaramu y'urubuga dukoresheje umuyoboro wa DevOps. Turashobora gukoresha DSPy kugirango twubake sisitemu yo gufata ibyemezo bya AI igena ingamba nziza zo kohereza zishingiye kubintu nkibikorwa bya porogaramu, ibitekerezo byabakoresha, hamwe nibikoresho bya sisitemu.

`` python kwinjiza dspy kuva muri langchain itumiza LLM

def yoherejwe_ibikorwa (porogaramu_imikorere, umukoresha_ibisubizo, sisitemu_isoko): # Koresha DSPy gutunganya amakuru no gufata ibyemezo amakuru = dspy.Data (porogaramu_imikorere, umukoresha_ibisubizo, sisitemu_isoko) icyemezo = dspy.Umufata Icyemezo (data)

# Koresha LangChain kugirango ubyare gahunda yo kohereza
llm = LLM ("inyandiko-davinci-003")
gahunda = llm.

gahunda yo kugaruka

Sobanura umuyoboro wa DevOps

def umuyoboro (porogaramu_imikorere, umukoresha_ibisubizo, sisitemu_isoko): # Koresha DSPy kugirango utegure umuyoboro umuyoboro = dspy. Umuyoboro () umuyoboro. umuyoboro.add_task (dspytasks. Kora, gahunda)

# Koresha umuyoboro
umuyoboro.run ()

Koresha umuyoboro

umuyoboro (90, 4.5, 80) `` `

Hejuru 5-10 Amabwiriza Yakoreshejwe Byinshi / Udukoryo

Dore top 5-10 zikoreshwa cyane / ibisobanuro kuri DSPy na LangChain:

1. dspy.Data (): Kurema ikintu cyamakuru yo gutunganya no gufata ibyemezo.
2. dspy. IcyemezoMaker (): Ifata ibyemezo bishingiye kumibare n'imiterere.
3. LLM.generate_text (): Yibyara inyandiko ishingiye kubitekerezo byinjira.
4. dspytasks.Deploy (): Kohereza porogaramu ishingiye kuri gahunda yo kohereza.
5. dspy. Umuyoboro (): Utegura umuyoboro wa DevOps.
6. umuyoboro.add_task (): Ongeraho inshingano kumuyoboro.
7. umuyoboro.run (): Koresha umuyoboro.
8. dspy.Data.process (): Itunganya amakuru yo gufata ibyemezo.
9. dspy.DecisionMaker.decide (): Ifata icyemezo gishingiye kumibare n'imiterere.
10. LLM.generate_text (prompt): Yibyara inyandiko ishingiye kubitekerezo byinjira.

Kubona amashusho

Hano hari amashusho ya agent ya AI ikoreshwa na DevOps umuyoboro ukoresheje Chart.js: json { "ubwoko": "akabari", "amakuru": { "ibirango": ["Gutunganya Data", "Gufata Ibyemezo", "Kohereza"], "imibare": [ { "label": "Agentic AI", "amakuru": [10, 20, 30], "backgroundColor": "rgba (255, 99, 132, 0.2)", "umupaka w'amabara": "rgba (255, 99, 132, 1)", "Ubugari bwumupaka": 1 } ] }, "amahitamo": { "umunzani": { "yAxes": [ { "amatiku": { "gutangiraAtZero": ukuri } } ] } } } ` Iyerekwa ryerekana ibyiciro bitatu byumushinga wa AI ukoreshwa na DevOps umuyoboro: gutunganya amakuru, gufata ibyemezo, no kohereza.

Umwanzuro

Muri iki kiganiro, twasesenguye igitekerezo cyo gutunganya AI agent mu miyoboro ya DevOps dukoresheje DSPy na LangChain. Twerekanye urugero rworoshye rwuburyo bwo gutangiza gahunda yo gukoresha urubuga dukoresheje sisitemu yo gufata ibyemezo bya AI. Twatanze kandi amashusho yumurongo wa AI ukoreshwa na DevOps umuyoboro ukoresheje Chart.js. Muguhuza AI ya agent mu miyoboro ya DevOps, amashyirahamwe arashobora kugera kuri ZeroOps no gukoresha software byihuse, hamwe nubwiza buhanitse, kandi ku giciro gito.