Deriv Emeghe Deriv One ka Ọ Bụrụ Open Source, Platfọm AI Ụlọọrụ Ya Nke Na-ebugharị Onwe Ya

30 Ọktoba 2026

Ihe atụ nke platfọm Deriv One: dashboard etiti jikọtara na onye inyeaka AI, ngalaba nka, ozi na ọrụ igwe ojii, na profaịlụ onye ọrụ

Dubai, 29 Ọktoba, 2026 – Deriv ewepụtala Deriv One, platfọm AI nke a na-ebugharị n'onwe ya, dị ka open source n'okpuru laịsensị MIT. Ọ na-eme ka onye ọbụla n'ụlọọrụ jiri nka AI nke ọrụ ha kwadoro, onye ọbụla na-emepụta sọftụwia nwekwara ike ịgba ya ọsọ. Ọ bụ n'efu, ịkọrọ akaụntụ onye na-ere ahịa adịghị mkpa.

Ohere maka onye ọbụla, na njikwa dị n'ime ya

Deriv One malitere dị ka arịrịọ Rakshit Choudhary, Onyeisi Ụlọọrụ (CEO), maka ndekọ nka. Ka agents na-agbasa, otu ọrụ ọbụla dere nka nke ya, ọ dịghịkwa onye nwere ike ịmata onye kwesịrị ịgba nka gịnị.

“AI ekwesịghị ịbụ ihe e kwenyere naanị ndị ole na ole nwere ike ịhazi agent nke ha. Site na Deriv One, onye ọbụla n'ụlọọrụ nwere ike iji nka nke ọrụ ha kwadoro, e nwekwara ihe nchebe (guardrails) a hazichaworị,” ka Rotimi Akinyele, onye a họpụtara ọhụrụ dị ka Onyeisi Teknụzụ (Chief Technology Officer) na Deriv, kwuru.

Mgbanwe nke AI bụ ihe mbụ abanyela n'ime Deriv nke ukwuu. N'okpuru ndu Choudhary, Deriv ekpebiela iji AI mezuo 75% nke ọrụ ndị a na-eme n'aka tupu ngwụcha afọ 2026. Deriv One, bụ ngalaba dị n'ime ụlọọrụ dị n'elu ụdị AI, bụ ebe ndị ọrụ na-agbanwe nka ha ka ọ bụrụ ọrụ agent. Nka 221 sitere n'aka ndị ọrụ 54 n'ụlọọrụ ahụ, ọ bụghị naanị otu otu.

Ihe onye ọbụla na-enweta

Ebe ọrụ: ngalaba kewara kewa, mmepụta dị ka taabụ, na nzikọrịta gaa ụdị AI ndị dị ọnụ ala.
Nka: akwụkwọ markdown e nyochara, nke a na-ahụ dabere na ọrụ onye.
Agents: akpaghị Slack nke ha, nke ọbụla nwere onye nwe aha ya.
Lab: kọwaa ngwa n'asụsụ dị mfe, lelee ya tupu ịgbasaa ya.

Diagram of Deriv One architecture: the web workspace and Slack feed four surfaces (Workspace, Lab, Skills, Agents), which sit on a shared core of identity and directory, secrets vault, sandboxed execution, skills catalogue, and usage and cost tracking, all self-hosted and connected to your own model gateway

Maka ndị na-emepụta sọftụwia: gbaa ya site n'iwu atọ

“Anyị na-ewepụta otu ebe ọrụ, katalọgụ nka, ndebanye aha agent na Lab nke anyị na-eji n'ime ụlọọrụ. Mee ya clone, duzie ya gaa na gateway ụdị AI nke gị, hụkwa etu ohere AI a na-achịkwa si arụ ọrụ n'ezie,” ka Akinyele kwuru.

Duzie ya gaa na gateway ọbụla kwekọrọ na OpenAI, gụnyere LM Studio maọbụ Ollama dị na mpaghara:

git clone [REPO_URL] && cd platform
npm run setup -- --quick --test-gateway
npm run dev:stack

Ọ gụnyere sandboxing, Docker, Kubernetes manifests na ntuziaka GKE na EKS. Plugins igwe ojii nke Deriv na-eji arụ ọrụ, njikọ ime ụlọọrụ na ụfọdụ atụmatụ Lab ewepụghị; ndị ga-eji ya ga-eweta gateway, onye na-enye njirimara (identity provider) na igwe ojii nke ha.

“Ụdị AI ga-anọgide na-agbanwe. Ihe siri ike bụ ngalaba dị n'okpuru ya: njirimara, ihe nzuzo, nyocha nka na sandboxing. Nke a bụ ihe anyị na-ekerịta, ka ndị otu ọzọ ghara ịmaliteghachi ya ọzọ,” ka Akinyele kwuru.

Njikọ ndị: GitHub · Laịsensị MIT · Atụmatụ nchekwa

Maka Deriv

Ihe karịrị afọ 26, Deriv na-arụsi ọrụ ike ime ka azụmahịa ịntanetị dị mfe inweta nye onye ọbụla, n'ebe ọbụla. N'ịbụ ụlọọrụ karịrị ndị ọrụ azụmahịa nde atọ n'ụwa niile tụkwasịrị obi, ụlọọrụ a na-enye ọtụtụ ụdị azụmahịa dị iche iche n'ahịa ndị a ma ama, site na platfọm azụmahịa ya ndị natara ihe nrite. Site na ndị ọrụ karịrị 900 n'obodo 13 n'ụwa niile, Deriv na-aga n'ihu ịrụsi ọrụ ike ka e nwee ihe ọmụma gbasara AI n'ọkwa ọbụla, na-ewulitekwa ọnọdụ ọrụ ebe ndị ọrụ ya nwere ike ito na-eso ụlọọrụ ahụ. Ka ụlọọrụ a na-eduzi mgbanwe nke AI dị ka isi ihe n'ọrụ ọbụla, ụkpụrụ ya nke iguzosi ike n'eziokwu, ikike, ịrụkọ ọrụ dị ka otu, na ilekwasị anya n'ahịa na-adịgide mgbe niile.

Onye a ga-akpọtụrụ maka akwụkwọ ozi: pr@deriv.com