##Chiyambi

Kusintha kofulumira kwa Artificial Intelligence (AI) ndi Machine Learning (ML) kwachititsa kuti agentic AI, mtundu wa AI womwe ungathe kuchitapo kanthu ndikupanga zisankho paokha. Ukadaulo uwu uli ndi kuthekera kosintha momwe timayendera kamangidwe ka mapulogalamu ndi makina a DevOps. M'nkhaniyi, tiwona kuphatikizika kwa agent AI, Large Language Models (LLMs), ndi zida za orchestration monga LangChain ndi DSPy kuti tigwiritse ntchito kamangidwe ka mapulogalamu, zomangamanga, ndi njira za DevOps.

Agent AI ndi chiyani?

Agency AI imatanthawuza mtundu wa AI womwe ungathe kuchitapo kanthu ndikupanga zisankho paokha. Izi ndizosiyana ndi machitidwe achikhalidwe a AI, omwe nthawi zambiri amapangidwa kuti azigwira ntchito zinazake ndikutsata malamulo omwe adafotokozedwatu. Machitidwe a Agetic AI, kumbali ina, amapangidwa kuti azikhala odzilamulira komanso osinthika, kuwalola kuphunzira kuchokera kumadera awo ndikupanga zisankho kutengera zolinga ndi zolinga zawo.

Udindo wa LLMs mu Agentik AI

Zinenero Zazinenelo Zazikulu (LLMs) ndi mtundu wa AI womwe umaphunzitsidwa pazambiri zambiri zamalemba ndipo amatha kupanga chilankhulo chonga cha anthu. Ma LLM ndi gawo lalikulu la machitidwe a AI, chifukwa amapereka luso lomvetsetsa ndikupanga chilankhulo cha anthu. Izi zimathandiza machitidwe a AI kuti azilankhulana ndi anthu ndi machitidwe ena mwachilengedwe komanso mwachilengedwe.

Zida Zoyimba: LangChain ndi DSPy

Zida za orchestration monga LangChain ndi DSPy zapangidwa kuti ziziyendetsa ndikugwirizanitsa machitidwe pakati pa machitidwe ndi zigawo zosiyanasiyana. Zidazi zimapereka njira yofotokozera ndi kuyendetsa kayendetsedwe ka ntchito, zomwe zingaphatikizepo ntchito monga kukonza deta, kuphunzitsa zitsanzo, ndi kutumiza.

LangChain

LangChain ndi chida chotsegulira gwero lotseguka chomwe chimapereka chiwongolero chosinthika komanso chokulirapo pomanga ndi kuyang'anira kayendedwe ka ntchito. LangChain idapangidwa kuti izigwira ntchito ndi machitidwe osiyanasiyana ndi zida, kuphatikiza ma LLM, ma database, ndi ma API.

DSPy

DSPy ndi chida cha Python cha orchestration chomwe chimapereka njira yosavuta komanso yodziwikiratu yofotokozera ndikuyendetsa kayendedwe ka ntchito. DSPy idapangidwa kuti izigwira ntchito ndi machitidwe ndi magawo osiyanasiyana, kuphatikiza ma LLM, databases, ndi ma API.

Makina Opanga Mapulogalamu a Mapulogalamu ndi Njira za DevOps

Agentic AI, LLMs, ndi zida zoyimba ngati LangChain ndi DSPy zitha kugwiritsidwa ntchito kupanga makina osiyanasiyana a mapulogalamu ndi njira za DevOps. Zitsanzo zina ndi izi:

Kuwunika kachidindo kodzichitira: Makina a Agetic AI atha kugwiritsidwa ntchito kuwunikanso ma code ndikuzindikira zomwe zingachitike, monga nsikidzi ndi zovuta zachitetezo.
Kuyesa paokha: Makina a Agetic AI atha kugwiritsidwa ntchito poyesa kuyesa, kuphatikiza kuyesa mayunitsi, kuyesa kuphatikiza, komanso kuyesa komaliza.
Kutumiza mokhazikika: Makina a Agentic AI atha kugwiritsidwa ntchito kuti azitha kutumizira, kuphatikiza kuyika ma code, makonzedwe, ndi zina.
Kuwunika kodzichitira: Makina a Agetic AI atha kugwiritsidwa ntchito kuwunikira, kuphatikiza kuyang'anira momwe makina amagwirira ntchito, chitetezo, ndi ma metrics ena.

Kukhazikitsa Kothandiza

M'chigawo chino, tipereka chitsanzo chosavuta cha momwe tingagwiritsire ntchito LangChain ndi DSPy kupanga makina opangira mapulogalamu ndi ndondomeko ya DevOps.

Chitsanzo: Ndemanga ya Khodi Yokhazikika

Muchitsanzo ichi, tidzagwiritsa ntchito LangChain ndi DSPy kuti tigwiritse ntchito ma code code. Tidzafotokozera ndondomeko ya ntchito yomwe ili ndi njira zotsatirazi:

Kusanthula kachidindo: Mayendedwe a ntchito adzagwiritsa ntchito LLM kusanthula kachidindo ndi kuzindikira zomwe zingachitike.
**Kubwereza ndondomeko **: Kuyenda kwa ntchito kudzagwiritsa ntchito dongosolo la AI la agent kuti liwonenso kachidindo ndi kuzindikira zomwe zingatheke.
** Kutumiza ma Code **: Kuyenda kwa ntchito kudzagwiritsa ntchito chida chotumizira kuti chitumize kachidindo kumalo opangira.

Nachi chitsanzo cha momwe mungagwiritsire ntchito kayendedwe ka ntchito pogwiritsa ntchito LangChain ndi DSPy:

import langchain
kuchokera ku langchain import LLM

# Tanthauzirani LLM
llm = LLM("text-davinci-003")

# Tanthauzirani kachitidwe ka AI
ai_system = langchain.AgenticAI()

# Sinthani magwiridwe antchito
ntchito = langchain.Workflow(
    masitepe=[
        langchain.Step(
            name="Code analysis",
            func=llm.analyze_code,
            inputs={"code": "def hello_world(): print('Moni World')"},
        ),
        langchain.Step(
            dzina = "Kubwereza Code",
            func=ai_system.review_code,
            inputs={"code": "def hello_world(): print('Moni World')"},
        ),
        langchain.Step(
            name="Code deployment",
            func=langchain.deploy_code,
            inputs={"code": "def hello_world(): print('Moni World')"},
        ),
    ]
)

# Pangani ndondomeko ya ntchito
workflow.execute ()

Khodi iyi imatanthawuza kayendetsedwe ka ntchito komwe kumaphatikizapo njira zitatu: kusanthula ma code, kubwereza ma code, ndi kutumiza ma code. Kuyenda kwa ntchito kumagwiritsa ntchito LLM kusanthula kachidindo, kachitidwe ka AI kachitidwe kuti awonenso kachidindo, ndi chida chotumizira kuti atumize kachidindo kumalo opangira.

Top 5-10 Omwe Amagwiritsidwa Ntchito Kwambiri / Zidutswa

Nawa malamulo/zidutswa 5-10 zomwe zimagwiritsidwa ntchito kwambiri pa LangChain ndi DSPy:

langchain.Workflow(masitepe=[...]): Tanthauzirani kayendedwe ka ntchito kokhala ndi masitepe angapo.
langchain.Step(name="...", func=..., inputs={...}): Tanthauzirani sitepe imodzi mumayendedwe a ntchito.
llm.analyze_code(inputs={"code": ...}): Gwiritsani ntchito LLM kusanthula khodi.
ai_system.review_code(inputs={"code": ...}): Gwiritsani ntchito makina a AI kuti muwunikenso khodi.
langchain.deploy_code(inputs={"code": ...}): Ikani khodi kumalo opangira.
langchain.AgenticAI(): Pangani chitsanzo cha agent AI system.
langchain.LLM(): Pangani chitsanzo cha LLM.
langchain.deploy_code(): Ikani khodi kumalo opangira.
langchain.execute_workflow(): Pangani ndondomeko ya ntchito.
langchain.get_workflow_status(): Pezani momwe kayendetsedwe ka ntchito.

# Kuwona

Nachi chitsanzo cha momwe mungawonere kayendedwe ka ntchito pogwiritsa ntchito Chart.js:

{
  "type": "bar",
  "data": {
    "labels": ["Code Analysis", "Code Review", "Code Deployment"],
    "ma dataset": [
      {
        "label": "Njira zogwirira ntchito",
        "data": [10, 20, 30],
        "backgroundColor": [
          "rgba(255, 99, 132, 0.2)",
          "rgba(54, 162, 235, 0.2)",
          "rgba(255, 206, 86, 0.2)"
        ],
        "BorderColor": [
          "rgba(255, 99, 132, 1)",
          "rgba(54, 162, 235, 1)",
          "rgba(255, 206, 86, 1)"
        ],
        "BorderWidth": 1
      }
    ]
  },
  "zosankha": {}
}

Khodi iyi imatanthawuza tchati cha bar chomwe chili ndi zilembo zitatu: Kusanthula Makhombo, Kubwereza Makhodi, ndi Kutumiza Makhombo. Tchati chikuwonetsa kuchuluka kwa masitepe pagawo lililonse la kayendetsedwe ka ntchito.

##Mapeto

Agentic AI, LLMs, ndi zida zoyimba ngati LangChain ndi DSPy ali ndi kuthekera kosintha momwe timayendera mamangidwe a mapulogalamu ndi makina a DevOps. Mwa kupanga ntchito monga kuwunika kwa ma code, kuyesa, ndi kutumiza, titha kupititsa patsogolo luso lathu lachitukuko. M'nkhaniyi, tafufuza kuphatikizidwa kwa matekinolojewa ndikupereka chitsanzo chosavuta cha momwe tingawagwiritsire ntchito kupanga makina opangira mapulogalamu ndi ndondomeko ya DevOps.

Leveraging Agent AI for Software Architecture ndi DevOps Automation