Mufananidzo uyu wakaitwa naArthur Kuzin), inopfupisa chaizvo zviri mukati meiyo blog post. Nekuda kweizvozvo, rondedzero inotevera inofanirwa kunzwisiswa zvakanyanya senyaya yeChishanu pane sechinhu chinonyanya kubatsira uye chehunyanzvi. Mukuwedzera, zvakakosha kuziva kuti zvinyorwa zvakapfuma mumashoko eChirungu. Handizivi kushandura mamwe acho nemazvo, uye handisi kuda kushandura mamwe acho.
Kuti shanduko kubva kunharaunda yedzidzo kuenda kunzvimbo yeindasitiri yakaitika sei inoburitswa muzvikamu zviviri zvekutanga. Mune ino, nhaurirano ichava pamusoro pezvakaitika zvinotevera.
Akanga ari January 2017. Panguva iyoyo, ndaiva nerinopfuura gore rechiitiko chebasa uye ndakashanda muSan Francisco mukambani saSr. Data Scientist.
TrueAccord chirongwa chekuunganidza chikwereti. Mumashoko akareruka - sangano rekuunganidza. Vaunganidzi vanowanzodaidza zvakawanda. Takatumira maemail akawanda, asi takafona mashoma. Imeyili yega yega yakatungamira kune webhusaiti yekambani, uko mukwereti akapihwa kuderedzwa pachikwereti, uye kunyange kubvumidzwa kubhadhara muzvikamu. Iyi nzira yakaunza kuunganidzwa kuri nani, kubvumidzwa kuwedzera uye kushomeka kuratidzwa kune mhosva.
The company was normal. Chigadzirwa chacho chakajeka. Management ine njere. Nzvimbo yakanaka.
Paavhareji, vanhu vari mumupata vanoshanda munzvimbo imwe kwerinenge gore nehafu. Ndiko kuti, chero kambani yaunoshandira ingori nhanho diki. Pane ino nhanho iwe unozokwidza imwe mari, tora ruzivo rutsva, hunyanzvi, kubatanidza uye mitsara mune yako tangazve. Mushure meizvi pane shanduko kune nhanho inotevera.
PaTrueAccord pachayo, ndakabatanidzwa mukubatanidza masisitimu ekurudziro kumatsamba eemail, pamwe nekuisa pamberi mafoni. Mhedzisiro inonzwisisika uye yakayerwa zvakanaka mumadhora kuburikidza neA/B kuyedzwa. Sezvo pakanga pasina kudzidza muchina ndisati ndasvika, tapuro yebasa rangu yakanga isina kuipa. Zvekare, zviri nyore kuvandudza chimwe chinhu pane chatove chakagadziridzwa zvakanyanya.
Mushure memwedzi mitanhatu yekushanda pane idzi masisitimu, vakatokwidza mubhadharo wangu kubva pamadhora zana nemakumi mashanu kusvika kumadhora 150k. Munharaunda pane meme inoita $163k. Inokura nemakumbo ayo kubva pano.
Zvose izvi zvaishamisa, asi hazvina kutungamirira chero kupi zvako, kana kuti zvakatungamirira, asi kwete ikoko.
Ndine ruremekedzo rukuru kune TrueAccord, zvose kambani nevakomana vandakashanda navo ikoko. Ndakadzidza zvakawanda kubva kwavari, asi ndakanga ndisingadi kushanda kwenguva yakareba pazvirongwa zvekukurudzira pane imwe sangano rekuunganidza. Kubva padanho iri waifanira kutsika mune imwe nzira. Kana zvisiri mberi uye kumusoro, saka zvishoma kumativi.
Chii chandisina kufarira?
- Kubva pakuona kwemichina yekudzidza, matambudziko haana kundifadza. Ini ndaida chimwe chinhu chine fashoni, chepwere, ndiko kuti, Kudzidza Kwakadzika, Kombuta Vision, chimwe chinhu chiri padyo nesainzi kana kuti alchemy.
- Kutanga, uye kunyange sangano rekuunganidza, rine matambudziko ekuhaya vashandi vane hunyanzvi. Sekutanga, haigone kubhadhara zvakanyanya. Asi sesangano rekuunganidza, rinorasikirwa nechimiro. Zvichireva kutaura, kana musikana ari kufambidzana achibvunza kwaunoshanda? Mhinduro yako: "PaGoogle" inonzwika maodha ehukuru zvirinani pane "sangano rekuunganidza." Ndakanetseka zvishoma nenyaya yekuti kune shamwari dzangu dzinoshanda kuGoogle neFacebook, kusiyana neni, zita rekambani yavo rakavhura mikova senge: unogona kukokwa kumusangano kana kusangana semukurukuri, kana vamwe vanhu vanofarira vanonyora pa LinkedIn. nechipo chekusangana uye kutaura pamusoro pegirazi reti. Ndinonyatsoda kukurukura nevanhu vandisingazive pachangu. Saka kana uchigara muSan Francisco, usazeze kunyora - handei kunonwa kofi titaure.
- Pamusoro pangu, vatatu veData Scientists vakashanda mukambani. Ini ndanga ndichishanda pamuchina kudzidza, uye ivo vaishanda pane mamwe maData Sayenzi mabasa, ayo akajairika mune chero kutanga kubva pano kusvika mangwana. Nekuda kweizvozvo, havana kunyatsonzwisisa kudzidza kwemichina. Asi kuti ndikure, ndinofanira kutaurirana nemumwe munhu, kukurukura nyaya uye zvichangobva kuitika, uye kukumbira mazano, pakupedzisira.
Chii chaivapo?
- Dzidzo: fizikisi, kwete sainzi yekombuta.
- Mutauro wepurogiramu chete wandaiziva waiva Python. Paive nemanzwiro ekuti ndaifanira kuchinjira kuC ++, asi ndakanga ndisingakwanise kusvika pazviri.
- Gore nehafu yebasa muindasitiri. Uyezve, kubasa handina kudzidza Deep Learning kana Computer Vision.
- Hapana kana chinyorwa chimwe chete paKudzidza Kwakadzika / Kombuta Vision mukutangazve.
- Paive neKaggle Master kubudirira.
Waida chii?
- Nzvimbo iyo ichave yakakosha kudzidzisa akawanda network, uye padyo nekombuta kuona.
- Zvirinani kana iri kambani hombe seGoogle, Tesla, Facebook, Uber, LinkedIn, nezvimwe. Kunyangwe mupinch, kutanga kwaizoita.
- Handidi kuti ndive nyanzvi huru yekudzidza muchina pachikwata. Kwakanga kuchidiwa zvikuru makomuredhi makuru, vaperekedzi uye marudzi ose ekurukurirano, izvo zvaifanira kukurumidzisa nzira yokudzidza.
- Mushure mekuverenga mablog ekuti vakapedza kudzidza vasina ruzivo rwemaindasitiri vane muripo wakakwana wemadhora mazana matatu-300k pagore, ndaida kupinda muchikamu chimwe chete. Hazvirevi kuti izvi zvinondishungurudza zvakanyanya, asi sezvo vachiti izvi zvinowanzoitika, asi ndine zvishoma, saka ichi chiratidzo.
Basa racho raiita serinogona kugadziriswa zvachose, kunyangwe zvisiri mupfungwa yekuti unogona kusvetukira mune chero kambani, asi kuti kana ukafa nenzara, zvese zvinobuda. Ndiko kuti, makumi kana mazana ekuedza, uye kurwadziwa kubva mukukundikana kwese uye kurambwa kwese, kunofanira kushandiswa kurodza kutarisa, kuvandudza chiyeuchidzo uye kutambanudza zuva kusvika kumaawa makumi matatu nematanhatu.
Ndakagadzirisa resume yangu, ndakatanga kuitumira kunze, uye kuenda kubvunzurudzo. Ndakabhururuka kupfuura vazhinji vavo padanho rekutaurirana naHR. Vanhu vazhinji vaida C ++, asi ini ndakanga ndisingazivi, uye ndakanga ndine manzwiro akasimba ekuti handingafariri zvikuru zvigaro zvinoda C ++.
Zvakakosha kucherechedza kuti panguva imwecheteyo pakanga paine shanduko yechikamu mumhando yemakwikwi paKaggle. Gore ra2017 risati rasvika pakanga paine data yakawanda yetabular uye kashoma data yemifananidzo, asi kutanga muna 2017 pakanga paine mabasa mazhinji ekuona komputa.
Hupenyu hwakayerera nenzira inotevera:
- Shanda masikati.
- Kana tech screen / onsite iwe unotora nguva kure.
- Manheru nekupera kwevhiki Kaggle + zvinyorwa / mabhuku / mabloggi
Kupera kwa2016 kwakaratidzwa nekuti ndakapinda munharaunda , iyo yakarerutsa zvinhu zvakawanda. Kune vakomana vakawanda munharaunda vane ruzivo rwakapfuma rwemabhizimisi, izvo zvakatibvumira kubvunza mibvunzo yakawanda yeupenzi uye kuwana mhinduro dzakawanda dzakachenjera. Kune zvakare akawanda akasimba kwazvo emuchina wekudzidza nyanzvi dzemitsetse yese, izvo, zvisingatarisirwi, zvakanditendera, kuburikidza neODS, kuvhara nyaya yacho nekutaurirana kwakadzama nguva dzose nezveData Science. Kusvika parizvino, maererano neML, ODS inondipa kakawanda kupfuura zvandinowana kubasa.
Zvakanaka, semazuva ese, ODS ine nyanzvi dzakakwana mumakwikwi paKaggle nedzimwe nzvimbo. Kugadzirisa matambudziko muchikwata kunonyanya kunakidza uye kubereka, saka nemajee, kutuka, memes uye kumwe kuvaraidzwa kwakashata, takatanga kugadzirisa matambudziko rimwe nerimwe.
Muna March 2017 - muboka neSerega Mushinsky - nzvimbo yechitatu ye . Menduru yegoridhe paKaggle + $20k yevaviri. Pabasa iri, shanda nemifananidzo yesatellite + binary segmentation kuburikidza neUNet yakagadziridzwa.
MunaKurume mumwe chete iwoyo, ndakaenda kunobvunzurudzwa kuNVidia neSelf Driving timu. Ndakanetseka chaizvo nemibvunzo nezve Object Detection. Pakanga pasina ruzivo rwakakwana.
Sezvineiwo, panguva imwecheteyo, makwikwi eObject Detection pamifananidzo yemuchadenga kubva kuDSTL imwechete yakatanga. Mwari pachake akarayira kugadzirisa dambudziko uye kuvandudza. Mwedzi wemanheru nekupera kwevhiki. Ndakanhonga zivo ndokupedza kechipiri. Makwikwi aya aive nekamwe kakunakidza mumitemo, zvakazoita kuti ndiratidzwe muRussia pane yemubatanidzwa kwete yemubatanidzwa. Ndakakwira , uye mune zvakawanda zvakadhindwa uye zvinyorwa zvepamhepo. Mail Ru Group yakagamuchira zvishoma zvakanaka PR pamari yangu nemari yayo, uye sainzi yakakosha muRussia yakapfumiswa ne12000 mapaundi. Semazuva ose, zvakanyorwa pamusoro penyaya iyi . Enda ikoko kuti uwane ruzivo.
Panguva imwecheteyo, Tesla recruiter akandibata ndokuda kutaura nezveComputer Vision chinzvimbo. Ndakabvuma. Ndakamhanya nekuenda kumba, maviri tekinoroji skrini, kubvunzurudzwa kwepanzvimbo, uye ndakava nehurukuro inonakidza naAndrei Karpathy, anga achangopihwa basa kuTesla seDirector weAI. Nhanho inotevera ndeyekutarisa kumashure. Mushure maizvozvo, Elon Musk aifanira kutendera chikumbiro changu. Tesla ine yakasimba Non Disclosure Agreement (NDA).
Handina kupasa cheki yebackgound. Anopinza vanhu basa akati ndataura zvakanyanya online, ndichityora iyo NDA. Nzvimbo chete yandakataura chero chinhu nezve kubvunzurudzwa kuTesla yaive ODS, saka fungidziro iripo ndeyekuti mumwe munhu akatora skrini akanyorera HR kuTesla, uye ini ndakabviswa mujaho kunze kwekukuvadza. Zvaitonyadzisa ipapo. Zvino ndinofara kuti hazvina kushanda. Nzvimbo yangu yazvino iri nani, kunyangwe zvingave zvinonakidza kushanda naAndrey.
Pakarepo mushure meizvozvo, ndakapinda mumakwikwi emifananidzo yesatellite paKaggle kubva . Dambudziko rakanga riri nyore uye raifinha zvakanyanya; hapana aida kurigadzirisa, asi munhu wese aida menduru yegoridhe yemahara kana mari yemubairo. Naizvozvo, neboka reKaggle Masters revanhu 7, takabvumirana kuti tichakanda simbi. Isu takadzidzisa 480 network mune 'fit_predict' modhi uye takagadzira nhanho-nhatu ensemble kubva mazviri. Takapedza wechinomwe. . Nenzira, Jeremy Howard, uyo anozivikanwa zvikuru semusiki akapedza 23.
Mushure mekupera kwemakwikwi, kuburikidza neshamwari yaishanda paAdRoll, ndakaronga Meetup panzvimbo yavo. Vamiriri vePlanet Labs vakataura ipapo nezve iyo sangano remakwikwi uye yekumaka data yakaita senge padivi pavo. Wendy Kwan, anoshanda kuKaggle uye akatarisa makwikwi, akataura nezvekuti akazviona sei. Ndakatsanangura mhinduro yedu, matekiniki, matekiniki uye ruzivo rwehunyanzvi. Zvikamu zviviri kubva muzvitatu zvevateereri vakagadzirisa dambudziko iri, saka mibvunzo yakabvunzwa kune imwe nzvimbo uye kazhinji zvinhu zvose zvakanga zvakanaka. Jeremy Howard aivepowo. Zvakazoitika kuti akapedza ari pachinzvimbo chechi23 nekuti aisaziva kurongedza modhi uye kuti aisaziva nezve nzira iyi yekugadzira ensembles zvachose.
Meetups mumupata pakudzidza muchina akasiyana zvakanyanya kubva kusangana muMoscow. Sezvo mutemo, kusangana mumupata ndiko pasi. Asi zvedu zvakava zvakanaka. Nehurombo, shamwari yaifanirwa kudzvanya bhatani uye kurekodha zvese haina kudzvanya bhatani :)
Mushure meizvozvo, ndakakokwa kuti nditaure kune chinzvimbo cheDeep Kudzidza Injiniya pane imwecheteyo Planet Labs, uye ipapo ipapo. handina kuzvipasa. Mazwi ekuramba ndeokuti hapana ruzivo rwakakwana muDeep Learning.
Ndakagadzira makwikwi ega ega seprojekiti mukati . Padambudziko reDSTL takanyora ndokuiisa pa arxiv. Kwete chinyorwa, asi zvakadaro chingwa. Iniwo ndinokurudzira kune vamwe vese kuti vawedzere yavo LinkedIn profile kuburikidza nemakwikwi, zvinyorwa, hunyanzvi, zvichingodaro. Pane kuwirirana kwakanaka pakati pekuti mangani mazwi akakosha aunawo mune yako LinkedIn profile uye kangani vanhu vanokutumira meseji.
Kana munguva yechando uye muchirimo ndaive nehunyanzvi, zvino muna Nyamavhuvhu ndaiva neruzivo uye kuzvivimba.
Pakupera kwaChikunguru, mumwe mukomana aishanda semaneja weData Science kuLyft akandibata paLinkedIn ndokundikoka kuti tinwe kofi uye titaure nezvehupenyu, nezve Lyft, nezve TrueAccord. Takataura. Akazvipira kubvunzurudza nechikwata chake pachinzvimbo cheData Scientist. Ndakati sarudzo iri kushanda, chero iri Computer Vision / Deep Kudzidza kubva mangwanani kusvika manheru. Akavimbisa kuti hapana chaaipikisa.
Ndakatumira resume yangu uye iye akaisa kuLyft's yemukati portal. Mushure maizvozvi, mushandirwi akandidana kuti ndivhure tsamba yangu uye ndizive zvakawanda pamusoro pangu. Kubva pamashoko ekutanga, zvaive pachena kuti kwaari ichi chaive chimiro, sezvo zvaive pachena kwaari kubva mukutanga kwake kuti "Ini handisi chinhu cheLyft." Ndofunga mushure meizvozvo resume yangu yakapinda mubhini remarara.
Nguva yese iyi, pandakanga ndichibvunzurudzwa, ndakakurukura nezvekutadza kwangu uye kutadza kwangu muODS uye vakomana vakandipa mhinduro uye vakandibatsira munzira dzese nemazano, kunyangwe, semazuva ese, pakanga painewo kutsika-tsika kwakawanda ikoko.
Mumwe wenhengo dzeODS akada kundibatanidza neshamwari yake, anova Director weEngineering kuLyft. Hapana kupera nguva. Ini ndinouya kuLyft kuzodya masikati, uye kunze kweshamwari iyi kune zvakare Musoro weData Sainzi uye Chigadzirwa maneja anove fan hombe yeDeep Kudzidza. Pakudya kwemasikati takakurukura pamusoro peDL. Uye sezvo ndanga ndichidzidzira network 24/7 kwehafu yegore, ndakaverenga cubic metres emabhuku, uye ndakamhanyisa mabasa paKaggle ine zvimwe kana zvishoma zvakajeka mhedzisiro, ndaigona kutaura nezve Kudzidza Kwakadzika kwemaawa, zvese maererano nezvinyorwa zvitsva uye. maitiro anoshanda .
Mushure mekudya kwemasikati vakanditarisa ndokuti - zviri pachena kuti wakanaka, unoda kutaura nesu here? Uyezve, ivo vakawedzera kuti zviri pachena kwandiri kuti tora kumba + tech skrini inogona kusvetuka. Uye kuti ini ndichakokwa nekukasira kune onsite. Ndakabvuma.
Pashure paizvozvo, anotora basa iyeye akandidana kuti ndironge kubvunzurudzwa panzvimbo, uye haana kugutsikana. Akataura chimwe chinhu pamusoro pekusasvetuka pamusoro pako.
Akauya. Onsite interview. Maawa mashanu ekutaurirana nevanhu vakasiyana. Pakanga pasina mubvunzo mumwechete nezve Kudzidza Kwakadzika, kana nezvekudzidza kwemuchina musimboti. Sezvo pasina Kudzidza Kwakadzika / Computer Vision, saka ini handisi kufarira. Nokudaro, mhinduro dzebvunzurudzo dzaive dzechokwadi.
Uyu muchengeti anofona ndokuti - makorokoto, wapfuura kubvunzurudzo yechipiri. Izvi zvose zvinoshamisa. Chii chechipiri chiri panzvimbo? Handisati ndambonzwa nezvechinhu chakadaro. Ndakaenda. Kune maawa mashoma ipapo, panguva ino zvese nezvechinyakare muchina kudzidza. Zviri nani. Asi zvakadaro hazvifadzi.
Munyoreri anofona nemakorokoto kuti ndakapasa bvunzurudzo yechitatu uye nemhiko kuti iyi ichave yekupedzisira. Ndakaenda kunoiona ndikaona iri DL neCV.
Ndakanga ndine wekutanga kwemwedzi yakawanda akandiudza kuti hapana chandaizopihwa. Ini ndichadzidzisa kwete pahunyanzvi hwehunyanzvi, asi pane dzakapfava. Kwete padivi rakapfava, asi pane chokwadi chekuti chinzvimbo chichavharwa kana kuti kambani haisati yatora basa, asi iri kungoyedza musika uye nhanho yevanoda.
Pakati paNyamavhuvhu. Ndakamwa beer okay. Mifungo yakasviba. 8 months dzapera and still no offer. Zvakanaka kugadzira pasi pedoro, kunyanya kana hunyanzvi hwacho huchishamisa. Imwe pfungwa inouya mumusoro mangu. Ndinoigovera naAlexey Shvets, uyo panguva iyoyo aiva postdoc paMIT.
Ko kana iwe ukatora iri padyo DL / CV musangano, tarisa makwikwi anoitwa sechikamu chayo, dzidzisa chimwe chinhu uye utumire? Sezvo vese nyanzvi dziripo dziri kuvaka mabasa avo pane izvi uye vanga vachiita izvi kwemwedzi yakawanda kana makore, isu hatina mukana. Asi hazvityisi. Isu tinoita kumwe kuzviisa pasi kunobatsira, kubhururuka kuenda kunzvimbo yekupedzisira, uye mushure mezvo tinonyora pre-kudhinda kana chinyorwa nezvekuti isu takasiyana sei nevamwe vese totaura nezvechisarudzo chedu. Uye chinyorwa chatova paLinkedIn uye mune yako kutangazve.
Ndiko kuti, zvinoita kunge zvine basa uye kune mamwe mazwi akakosha mukutangazve, izvo zvinofanirwa kuwedzera zvishoma mikana yekusvika kune tech skrini. Kodhi uye kutumira kubva kwandiri, zvinyorwa kubva kuAlexey. Game, hongu, asi sei kwete?
Hapana kupera nguva. Musangano wepedyo watakatsvaga pagoogle waive MICCAI uye paitove nemakwikwi ipapo. Takarova wekutanga. Zvanga zviri . Basa racho rine 3 subtasks. Kwakanga kwasara mazuva masere kuti nguva yekupedzisira isati yasvika. Ndakamuka mangwanani, asi handina kurega pfungwa yacho. Ndakatora mapaipi angu kubva kuKaggle ndokuachinja kubva kusatellite data kuenda kune yekurapa. 'fit_predict'. Alexey akagadzirira tsananguro yemapeji maviri emhinduro dzedambudziko rega rega, uye isu takaitumira. Ready. Muchidzidzo, unogona kufema. Asi zvakazoitika kuti pakanga paine rimwe basa remusangano mumwe chete () aine madiki matatu uye kuti nguva yake yekupedzisira yakakwidzwa nemazuva mana, kureva kuti, tinogona kuita 'fit_predict' ipapo nekuitumira. Ndozvatakaita.
Kusiyana neKaggle, makwikwi aya aive neavo ezvidzidzo zvehunyanzvi:
- Hapana Leaderboard. Zvikumbiro zvinotumirwa neemail.
- Iwe uchabviswa kana mumiriri wechikwata akasauya kuzopa mhinduro pamusangano paWorkshop.
- Nzvimbo yako pabhodhi rekutungamira inozozivikanwa chete panguva yemusangano. Imwe mhando yemutambo wezvidzidzo.
Musangano weMICCAI 2017 wakaitirwa muQuebec City. Kutaura chokwadi, munaGunyana ndakanga ndava kutanga kupera simba, saka pfungwa yekutora vhiki kubva kubasa uye kuenda kuCanada yairatidza kunakidza.
Ndakauya kumusangano. Ndauya kuWorkshop uku, hapana wandinoziva, ndakagara mukona. Wese munhu anoziva mumwe nemumwe, vanotaurirana, vanokanda kunze mazwi ekurapa akangwara. Ongororo yemakwikwi ekutanga. Vatori vechikamu vanotaura uye vanotaura nezvesarudzo dzavo. Ikoko kunotonhorera, nekupenya. Nguva yangu. Uye ini neimwe nzira ndinonyara. Vakagadzirisa dambudziko, vakashanda pariri, sainzi yepamusoro, uye isu tiri "fit_predict" kubva pane zvakaitika kare, kwete zvesainzi, asi kuti tisimudzire kutanga kwedu.
Akabuda ndokuti ini handisi nyanzvi mune zvekurapa, vakakumbira ruregerero nekutambisa nguva yavo, uye vakandiratidza imwe siraidhi ine mhinduro. Ndakadzika kuhoro.
Vanozivisa yekutanga subtask - isu tiri vekutanga, uye nemuganho.
Yechipiri neyechitatu inoziviswa.
Vanozivisa chechitatu - zvakare kutanga uye zvakare nekutungamira.
General ndiye wekutanga.
Vamwe muvateereri vanonyemwerera vakanditarisa neruremekedzo. Vamwe, avo sezviri pachena vairangarirwa kuva nyanzvi mubasa iri, vakanga vahwina mari yebasa iri uye vakanga vaita ikoku kwamakore akawanda, vakanga vane chimiro chakamonyaniswa zvishomanene pazviso zvavo.
Rinotevera ibasa rechipiri, iro rine matatu subtasks uye rakafambiswa mberi nemazuva mana.
Apa ndakakumbirawo ruregerero ndokuratidza slide yedu imwe zvakare.
Nyaya imwe chete. Zviviri zvekutanga, sekondi imwe, zvakajairika kutanga.
Ndinofunga kuti iyi ingangove nguva yekutanga munhoroondo kuti sangano rekuunganidza rakunda mukwikwidzi wekufungidzira wezvekurapa.
Uye zvino ndakamira pachikuva, vari kundipa imwe mhando yedhipuroma uye ndinoputirwa. Iyo fuck ingave sei iyo? Vadzidzi ava vari kushandisa mari yevateresi, vachishanda kurerutsa uye kuvandudza kunaka kwebasa revanachiremba, kureva kuti, mukufungidzira, tarisiro yehupenyu hwangu, uye imwe mutumbi yakabvarura iyi vashandi vese vedzidzo mumureza weBritish manheru mashoma.
Bhonasi kune izvi nderekuti mune zvimwe zvikwata, vadzidzi vakapedza kudzidza vanga vachishanda pamabasa aya kwemwedzi yakawanda vachange vaine resume inoyevedza kuna HR, ndiko kuti, vanosvika nyore kune tech skrini. Uye pamberi pemaziso angu pane email ichangobva kugamuchirwa:
A Googler recently referred you for the Research Scientist, Google Brain (United States) role. We carefully reviewed your background and experience and decided not to proceed with your application at this time.Kazhinji, kubva pachikuva, ndinobvunza vateereri kuti: βPane anoziva kwandinoshanda here?β Mumwe wevarongi vemakwikwi aiziva - iye Googled kuti TrueAccord chii. Vamwe vose havasi. Ndinoenderera mberi: βNdinoshandira kambani inounganidza zvinhu, uye kubasa handiite Computer Vision kana Deep Learning. Uye nenzira dzakawanda, izvi zvinoitika nekuti madhipatimendi eHR eGoogle Brain uye Deepmind anosefa kutanga kwangu, kwete kundipa mukana wekuratidza kudzidziswa kwehunyanzvi. "
Vakatambidza chitupa, kuzorora. Boka revadzidzi vanondidhonzera parutivi. Zvakazoitika kuti iri iboka reHealth rine Deepmind. Vakafadzwa zvikuru zvekuti vakabva vada kutaura neni nezvenzvimbo yeResearch Engineer muchikwata chavo. (Takataura. Iyi hurukuro yakagara kwemwedzi ye6, ndakapfuura take home, quiz, asi yakagurwa pfupi pane tech screen. Mwedzi 6 kubva pakutanga kwekukurukurirana kusvika kune tech screen inguva yakareba. Kumirira kwenguva refu kunopa kuravira. Research Engineer at Deepmind muLondon, pakatarisana nechekumashure kweTrueAccord paive nedanho rakasimba rekukwira, asi kuseri kwechinzvimbo changu ikozvino idanho rekudzika.Kubva chinhambwe chemakore maviri apfuura kubva ipapo, zvakanaka. kuti handizvo.)
mhedziso
Panenge panguva imwe cheteyo, ndakagamuchira chikumbiro kubva kuLyft, icho ndakabvuma.
Zvichienderana nemhedzisiro yemakwikwi maviri aya neMICCAI, zvinotevera zvakaburitswa:
Ndokunge, kunyangwe kupenga kweiyo pfungwa, kuwedzera zvinyorwa zvekuwedzera uye preprints kuburikidza nemakwikwi zvinoshanda nemazvo. Uye mumakore akatevera takatowedzera kuipa.
Ndanga ndichishanda kuLyft kwemakore akati wandei apfuura ndichiita Computer Vision/Deep Learning for Self Driving mota. Kureva kuti ndawana zvandaida. Uye mabasa, nekambani yepamusoro-soro, uye vashandi vakasimba, nezvimwe zvose zvinonaka.
Kwemwedzi iyi, ndakave nekutaurirana nemakambani makuru maviri Google, Facebook, Uber, LinkedIn, uye negungwa rekutanga kwehukuru hwakasiyana.
Zvairwadza mwedzi yese iyi. Zvadenga zvinokuudza chimwe chinhu chisingafadzi mazuva ese. Kugara uchirambwa, kugara uchiita zvikanganiso uye zvese izvi zvinovhenekerwa nekuramba uchinzwa kwekusava netariro. Iko hakuna vimbiso yekuti uchabudirira, asi pane manzwiro ekuti uri benzi. Zvinondiyeuchidza kuti ndakaedza sei kuwana basa ndichangopedza yunivhesiti.
Ndinofunga kuti vakawanda vaitsvaka basa mumupata uye zvose zvakanga zviri nyore kwavari. Manomano, mukuona kwangu, ndeichi. Kana uri kutsvaga basa mumunda waunonzwisisa, une ruzivo rwakawanda, uye resume yako inotaura zvakafanana, hapana matambudziko. Ndakaitora ndokuiwana. Kune vacancies vakawanda.
Asi kana iwe uchitsvaga basa mumunda mutsva kwauri, ndiko kuti, kana pasina ruzivo, pasina hukama uye resume yako inotaura chimwe chinhu chakaipa - panguva ino zvese zvinova zvinonakidza.
Parizvino, vanyoreri vanogara vachindinyorera uye vanopa kuita chinhu chimwe chete chandiri kuita iye zvino, asi mune imwe kambani. Inguva chaiyo yekuchinja mabasa. Asi hapana chikonzero chekuti ndiite zvandinenge ndagona. Yechii?
Asi pane zvandinoda, ini zvakare handina ruzivo kana mitsetse mukutanga kwangu. Ngationei kuti zvese izvi zvinopera sei. Kana zvinhu zvose zvikafamba zvakanaka, ndichanyora chikamu chinotevera. π
Source: www.habr.com