{"id":1613,"date":"2026-04-21T03:57:51","date_gmt":"2026-04-21T03:57:51","guid":{"rendered":"https:\/\/bbotech.vn\/?p=1613"},"modified":"2026-04-21T07:14:13","modified_gmt":"2026-04-21T07:14:13","slug":"chatbot-sang-agent","status":"publish","type":"post","link":"https:\/\/bbotech.vn\/vi\/chatbot-sang-agent\/","title":{"rendered":"T\u1eeb Chatbot Sang AI Agent: V\u00ec Sao Doanh Nghi\u1ec7p C\u1ea7n Chuy\u1ec3n Tr\u1ecdng T\u00e2m Ngay T\u1eeb B\u00e2y Gi\u1edd"},"content":{"rendered":"

<\/p>\n

Doanh nghi\u1ec7p \u0111ang r\u1eddi kh\u1ecfi l\u1edbp AI ch\u1ec9 tr\u1ea3 l\u1eddi \u0111\u1ec3 ti\u1ebfn sang AI c\u00f3 th\u1ec3 th\u1ef1c thi c\u00f4ng vi\u1ec7c<\/h2>\n

N\u1ebfu n\u0103m 2024 l\u00e0 giai \u0111o\u1ea1n nhi\u1ec1u doanh nghi\u1ec7p th\u1eed chatbot \u0111\u1ec3 tr\u1ea3 l\u1eddi c\u00e2u h\u1ecfi, so\u1ea1n n\u1ed9i dung, ho\u1eb7c h\u1ed7 tr\u1ee3 nh\u00e2n vi\u00ean tra c\u1ee9u nhanh, th\u00ec \u0111\u1ebfn n\u0103m 2026 tr\u1ecdng t\u00e2m \u0111\u00e3 \u0111\u1ed5i. C\u00e2u h\u1ecfi kh\u00f4ng c\u00f2n l\u00e0 \u201cAI tr\u1ea3 l\u1eddi c\u00f3 hay kh\u00f4ng\u201d, m\u00e0 l\u00e0 \u201cAI c\u00f3 th\u1ec3 ho\u00e0n th\u00e0nh m\u1ed9t quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow) th\u1eadt, c\u00f3 ki\u1ec3m so\u00e1t, c\u00f3 \u0111o hi\u1ec7u qu\u1ea3 hay kh\u00f4ng\u201d.<\/p>\n

T\u00edn hi\u1ec7u th\u1ecb tr\u01b0\u1eddng \u0111\u00e3 kh\u00e1 r\u00f5. Gartner c\u00f4ng b\u1ed1 ng\u00e0y 26\/8\/2025<\/strong> r\u1eb1ng 40% \u1ee9ng d\u1ee5ng doanh nghi\u1ec7p s\u1ebd t\u00edch h\u1ee3p AI agent theo nhi\u1ec7m v\u1ee5 v\u00e0o cu\u1ed1i n\u0103m 2026<\/strong>, t\u0103ng t\u1eeb d\u01b0\u1edbi 5%<\/strong> trong n\u0103m 2025. (Ngu\u1ed3n: Gartner, August 2025)<\/em> C\u0169ng theo Gartner, \u0111\u1ebfn n\u0103m 2028<\/strong>, h\u01a1n m\u1ed9t n\u1eeda doanh nghi\u1ec7p<\/strong> s\u1ebd ng\u1eebng tr\u1ea3 ti\u1ec1n cho l\u1edbp AI mang t\u00ednh h\u1ed7 tr\u1ee3 nh\u01b0 tr\u1ee3 l\u00fd \u0111\u1ed3ng h\u00e0nh (copilot) hay c\u1ed1 v\u1ea5n th\u00f4ng minh (smart advisor), \u0111\u1ec3 chuy\u1ec3n sang n\u1ec1n t\u1ea3ng cam k\u1ebft k\u1ebft qu\u1ea3 quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow). (Ngu\u1ed3n: Gartner, April 2026)<\/em><\/p>\n

\u0110i\u1ec1u \u0111\u00f3 c\u00f3 ngh\u0129a l\u00e0 giai \u0111o\u1ea1n \u201cc\u00f3 chatbot l\u00e0 \u0111\u1ee7\u201d \u0111ang k\u1ebft th\u00fac nhanh h\u01a1n nhi\u1ec1u \u0111\u1ed9i ng\u0169 d\u1ef1 \u0111o\u00e1n. V\u1edbi CTO, QA Lead hay Engineering Manager, \u0111\u00e2y kh\u00f4ng ch\u1ec9 l\u00e0 c\u00e2u chuy\u1ec7n c\u00f4ng c\u1ee5 m\u1edbi. \u0110\u00e2y l\u00e0 thay \u0111\u1ed5i v\u1ec1 ki\u1ebfn tr\u00fac v\u1eadn h\u00e0nh: t\u1eeb AI \u0111\u1ee9ng ngo\u00e0i quy tr\u00ecnh sang AI \u0111\u01b0\u1ee3c trao quy\u1ec1n x\u1eed l\u00fd m\u1ed9t ph\u1ea7n c\u00f4ng vi\u1ec7c trong ranh gi\u1edbi ch\u00ednh s\u00e1ch (policy) r\u00f5 r\u00e0ng.<\/p>\n

\u0110i\u1ec3m chuy\u1ec3n pha quan tr\u1ecdng kh\u00f4ng n\u1eb1m \u1edf vi\u1ec7c AI n\u00f3i t\u1ef1 nhi\u00ean h\u01a1n, m\u00e0 \u1edf vi\u1ec7c AI \u0111\u00e3 b\u1eaft \u0111\u1ea7u \u0111\u01b0\u1ee3c giao quy\u1ec1n h\u00e0nh \u0111\u1ed9ng trong h\u1ec7 th\u1ed1ng doanh nghi\u1ec7p.<\/p><\/blockquote>\n

\"\"<\/p>\n

 <\/p>\n

AI agent l\u00e0 g\u00ec, v\u00e0 kh\u00e1c chatbot doanh nghi\u1ec7p \u1edf \u0111i\u1ec3m n\u00e0o?<\/h2>\n

AI agent<\/strong> l\u00e0 h\u1ec7 th\u1ed1ng kh\u00f4ng ch\u1ec9 tr\u1ea3 l\u1eddi c\u00e2u l\u1ec7nh (prompt), m\u00e0 c\u00f2n c\u00f3 th\u1ec3 nh\u1eadn m\u1ee5c ti\u00eau, t\u1ef1 chia nh\u1ecf c\u00f4ng vi\u1ec7c, g\u1ecdi c\u00f4ng c\u1ee5 ho\u1eb7c giao di\u1ec7n l\u1eadp tr\u00ecnh \u1ee9ng d\u1ee5ng (API), \u0111\u1ecdc tr\u1ea1ng th\u00e1i h\u1ec7 th\u1ed1ng, r\u1ed3i ho\u00e0n th\u00e0nh nhi\u1ec1u b\u01b0\u1edbc li\u00ean ti\u1ebfp trong gi\u1edbi h\u1ea1n \u0111\u01b0\u1ee3c thi\u1ebft k\u1ebf s\u1eb5n. Chatbot t\u1ed1i \u01b0u cho h\u1ed9i tho\u1ea1i; agent t\u1ed1i \u01b0u cho k\u1ebft qu\u1ea3.<\/p>\n

Kh\u00e1c bi\u1ec7t n\u00e0y nghe c\u00f3 v\u1ebb nh\u1ecf tr\u00ean slide, nh\u01b0ng r\u1ea5t l\u1edbn khi \u0111i v\u00e0o v\u1eadn h\u00e0nh. M\u1ed9t chatbot ch\u0103m s\u00f3c kh\u00e1ch h\u00e0ng c\u00f3 th\u1ec3 g\u1ee3i \u00fd c\u00e2u tr\u1ea3 l\u1eddi cho nh\u00e2n vi\u00ean. M\u1ed9t agent ch\u0103m s\u00f3c kh\u00e1ch h\u00e0ng c\u00f3 th\u1ec3 t\u1ef1 ph\u00e2n lo\u1ea1i phi\u1ebfu x\u1eed l\u00fd (ticket), tra c\u1ee9u l\u1ecbch s\u1eed giao d\u1ecbch, \u0111\u1ec1 xu\u1ea5t ph\u01b0\u01a1ng \u00e1n x\u1eed l\u00fd, \u0111\u1ea9y t\u00ecnh hu\u1ed1ng (case) sang \u0111\u00fang h\u00e0ng ch\u1edd, v\u00e0 ch\u1ec9 chuy\u1ec3n c\u1ea5p (escalate) sang ng\u01b0\u1eddi th\u1eadt khi ch\u1ea1m ng\u01b0\u1ee1ng r\u1ee7i ro. Trong k\u1ef9 thu\u1eadt ph\u1ea7n m\u1ec1m c\u0169ng v\u1eady: chatbot gi\u00fap vi\u1ebft tr\u01b0\u1eddng h\u1ee3p ki\u1ec3m th\u1eed (test case) nhanh h\u01a1n; agent c\u00f3 th\u1ec3 b\u00e1m ch\u00eanh l\u1ec7ch m\u00e3 ngu\u1ed3n (code diff), ch\u1ecdn b\u1ed9 ki\u1ec3m th\u1eed h\u1ed3i quy (regression suite) li\u00ean quan, m\u1edf phi\u1ebfu l\u1ed7i (bug ticket) v\u00e0 c\u1eadp nh\u1eadt tr\u1ea1ng th\u00e1i sau khi ch\u1ea1y.<\/p>\n

N\u00f3i ng\u1eafn g\u1ecdn, chatbot tr\u1ea3 l\u1ea1i n\u1ed9i dung cho con ng\u01b0\u1eddi quy\u1ebft \u0111\u1ecbnh ti\u1ebfp. Agent b\u1eaft \u0111\u1ea7u tham gia v\u00e0o v\u00f2ng l\u1eb7p quy\u1ebft \u0111\u1ecbnh v\u00e0 th\u1ef1c thi.<\/p>\n

 <\/p>\n

<\/p>\n

Ba t\u00edn hi\u1ec7u cho th\u1ea5y th\u1ecb tr\u01b0\u1eddng \u0111ang chuy\u1ec3n t\u1eeb \u201cAI h\u1ed7 tr\u1ee3 (assistive AI)\u201d sang \u201cAI h\u01b0\u1edbng k\u1ebft qu\u1ea3 (outcome-focused AI)\u201d<\/h2>\n

1. L\u00e3nh \u0111\u1ea1o \u0111\u00e3 xem 2025-2026 l\u00e0 giai \u0111o\u1ea1n ph\u1ea3i thi\u1ebft k\u1ebf l\u1ea1i c\u00e1ch l\u00e0m vi\u1ec7c<\/h3>\n

Trong b\u00e1o c\u00e1o Work Trend Index 2025<\/em> c\u00f4ng b\u1ed1 ng\u00e0y 23\/4\/2025<\/strong>, Microsoft cho bi\u1ebft 82% l\u00e3nh \u0111\u1ea1o<\/strong> xem \u0111\u00e2y l\u00e0 n\u0103m b\u1ea3n l\u1ec1 \u0111\u1ec3 suy ngh\u0129 l\u1ea1i chi\u1ebfn l\u01b0\u1ee3c v\u00e0 v\u1eadn h\u00e0nh, c\u00f2n 81%<\/strong> k\u1ef3 v\u1ecdng agent s\u1ebd \u0111\u01b0\u1ee3c t\u00edch h\u1ee3p \u1edf m\u1ee9c v\u1eeba ho\u1eb7c l\u1edbn v\u00e0o chi\u1ebfn l\u01b0\u1ee3c AI c\u1ee7a c\u00f4ng ty trong 12-18 th\u00e1ng<\/strong>. (Ngu\u1ed3n: Microsoft Work Trend Index 2025)<\/em><\/p>\n

S\u1ed1 li\u1ec7u n\u00e0y quan tr\u1ecdng v\u00ec n\u00f3 cho th\u1ea5y th\u1ecb tr\u01b0\u1eddng kh\u00f4ng c\u00f2n xem agent l\u00e0 m\u1ed9t th\u1eed nghi\u1ec7m r\u00eca. N\u00f3 \u0111ang \u0111i v\u00e0o c\u00e2u chuy\u1ec7n t\u1ed5 ch\u1ee9c, quy m\u00f4 nh\u00e2n s\u1ef1 (headcount), ph\u00e2n vai gi\u1eefa con ng\u01b0\u1eddi v\u00e0 ph\u1ea7n m\u1ec1m. C\u0169ng trong nghi\u00ean c\u1ee9u n\u00e0y, 24% l\u00e3nh \u0111\u1ea1o<\/strong> n\u00f3i c\u00f4ng ty c\u1ee7a h\u1ecd \u0111\u00e3 tri\u1ec3n khai AI tr\u00ean to\u00e0n t\u1ed5 ch\u1ee9c, trong khi ch\u1ec9 12%<\/strong> v\u1eabn d\u1eebng \u1edf m\u1ee9c th\u00ed \u0111i\u1ec3m (pilot). (Ngu\u1ed3n: Microsoft Work Trend Index 2025)<\/em><\/p>\n

2. Doanh nghi\u1ec7p b\u1eaft \u0111\u1ea7u d\u00f9ng agent \u0111\u1ec3 t\u1ef1 \u0111\u1ed9ng h\u00f3a quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow), kh\u00f4ng ch\u1ec9 h\u1ed7 tr\u1ee3 t\u1eebng c\u00e1 nh\u00e2n<\/h3>\n

Microsoft ghi nh\u1eadn 46% l\u00e3nh \u0111\u1ea1o<\/strong> n\u00f3i doanh nghi\u1ec7p c\u1ee7a h\u1ecd \u0111ang d\u00f9ng agent \u0111\u1ec3 t\u1ef1 \u0111\u1ed9ng h\u00f3a ho\u00e0n to\u00e0n quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow) ho\u1eb7c quy tr\u00ecnh<\/strong>. (Ngu\u1ed3n: Microsoft Work Trend Index 2025)<\/em> \u0110\u00e2y l\u00e0 m\u1ed1c kh\u00e1c bi\u1ec7t gi\u1eefa chatbot v\u00e0 agent. Chatbot n\u00e2ng n\u0103ng su\u1ea5t c\u00e1 nh\u00e2n. Agent ch\u1ea1m v\u00e0o l\u1edbp quy tr\u00ecnh (process layer), n\u01a1i l\u1ee3i t\u1ee9c \u0111\u1ea7u t\u01b0 (ROI) th\u01b0\u1eddng l\u1edbn h\u01a1n nh\u01b0ng c\u0169ng \u0111\u00f2i h\u1ecfi qu\u1ea3n tr\u1ecb (governance) ch\u1eb7t h\u01a1n.<\/p>\n

Gartner c\u0169ng m\u00f4 t\u1ea3 r\u1ea5t r\u00f5 h\u01b0\u1edbng \u0111i \u0111\u00f3 trong th\u00f4ng c\u00e1o ng\u00e0y 2\/4\/2026<\/strong>: doanh nghi\u1ec7p s\u1ebd chuy\u1ec3n chi ti\u00eau t\u1eeb AI h\u1ed7 tr\u1ee3 sang AI cam k\u1ebft k\u1ebft qu\u1ea3 quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow). \u0110\u00e2y l\u00e0 m\u1ed9t thay \u0111\u1ed5i v\u1ec1 ng\u00e2n s\u00e1ch v\u00e0 c\u00e1ch mua ph\u1ea7n m\u1ec1m, kh\u00f4ng ch\u1ec9 l\u00e0 \u0111\u1ed5i giao di\u1ec7n t\u1eeb chat sang c\u00f4ng c\u1ee5 d\u1ef1ng quy tr\u00ecnh kh\u00f4ng c\u1ea7n m\u00e3 (no-code builder).<\/p>\n

3. T\u1ed1c \u0111\u1ed9 tri\u1ec3n khai \u0111ang t\u0103ng nhanh h\u01a1n t\u1ed1c \u0111\u1ed9 ho\u00e0n thi\u1ec7n governance<\/h3>\n

KPMG trong Q1 2026 AI Quarterly Pulse<\/em> cho bi\u1ebft 54% t\u1ed5 ch\u1ee9c<\/strong> \u0111\u00e3 ch\u1ee7 \u0111\u1ed9ng tri\u1ec3n khai AI agent trong v\u1eadn h\u00e0nh c\u1ed1t l\u00f5i (core operations). Nh\u01b0ng c\u00f9ng l\u00fac, 65% l\u00e3nh \u0111\u1ea1o<\/strong> n\u00f3i tr\u1edf ng\u1ea1i l\u1edbn nh\u1ea5t \u0111\u1ec3 ch\u1ee9ng minh l\u1ee3i t\u1ee9c \u0111\u1ea7u t\u01b0 (ROI) l\u00e0 m\u1edf r\u1ed9ng tr\u01b0\u1eddng h\u1ee3p s\u1eed d\u1ee5ng (scale use case), 62%<\/strong> nh\u1eafc t\u1edbi kho\u1ea3ng tr\u1ed1ng k\u1ef9 n\u0103ng, v\u00e0 91%<\/strong> cho r\u1eb1ng b\u1ea3o m\u1eadt, quy\u1ec1n ri\u00eang t\u01b0 v\u00e0 qu\u1ea3n tr\u1ecb r\u1ee7i ro s\u1ebd l\u00e0 ph\u1ea7n kh\u00f4ng th\u1ec3 t\u00e1ch kh\u1ecfi chi\u1ebfn l\u01b0\u1ee3c AI trong 6 th\u00e1ng t\u1edbi. (Ngu\u1ed3n: KPMG, Q1 2026)<\/em><\/p>\n

N\u00f3i c\u00e1ch kh\u00e1c, doanh nghi\u1ec7p kh\u00f4ng thi\u1ebfu h\u1ee9ng th\u00fa. H\u1ecd thi\u1ebfu m\u00f4 h\u00ecnh v\u1eadn h\u00e0nh \u0111\u1ee7 tr\u01b0\u1edfng th\u00e0nh \u0111\u1ec3 \u0111\u01b0a agent v\u00e0o quy tr\u00ecnh quan tr\u1ecdng m\u00e0 kh\u00f4ng t\u1ea1o th\u00eam n\u1ee3 k\u1ef9 thu\u1eadt, n\u1ee3 b\u1ea3o m\u1eadt v\u00e0 n\u1ee3 ki\u1ec3m so\u00e1t.<\/p>\n

 <\/p>\n

V\u00ec sao chatbot doanh nghi\u1ec7p b\u1eaft \u0111\u1ea7u ch\u1ea1m tr\u1ea7n gi\u00e1 tr\u1ecb<\/h2>\n

Chatbot v\u1eabn h\u1eefu \u00edch. N\u00f3 t\u1ed1t cho Q&A n\u1ed9i b\u1ed9, support c\u1ea5p m\u1ed9t, g\u1ee3i \u00fd n\u1ed9i dung, ho\u1eb7c t\u00ecm ki\u1ebfm tri th\u1ee9c doanh nghi\u1ec7p. V\u1ea5n \u0111\u1ec1 l\u00e0 nhi\u1ec1u doanh nghi\u1ec7p \u0111ang k\u1ef3 v\u1ecdng v\u00e0o chatbot m\u1ed9t lo\u1ea1i ROI m\u00e0 b\u1ea3n th\u00e2n ki\u1ebfn tr\u00fac chatbot kh\u00f3 t\u1ea1o ra.<\/p>\n

    \n
  • Chatbot th\u01b0\u1eddng d\u1eebng \u1edf l\u1edbp h\u1ed9i tho\u1ea1i:<\/strong> n\u00f3 tr\u1ea3 l\u1eddi t\u1ed1t nh\u01b0ng kh\u00f4ng t\u1ef1 x\u1eed l\u00fd c\u00f4ng vi\u1ec7c xuy\u00ean h\u1ec7 th\u1ed1ng.<\/li>\n
  • Chatbot \u00edt s\u1edf h\u1eefu ng\u1eef c\u1ea3nh th\u1ef1c thi:<\/strong> n\u00f3 c\u00f3 th\u1ec3 bi\u1ebft c\u00e2u tr\u1ea3 l\u1eddi \u0111\u00fang, nh\u01b0ng kh\u00f4ng c\u00f3 quy\u1ec1n, kh\u00f4ng c\u00f3 ch\u00ednh s\u00e1ch (policy) v\u00e0 kh\u00f4ng c\u00f3 nh\u1eadt k\u00fd ki\u1ec3m v\u1ebft (audit trail) \u0111\u1ec3 h\u00e0nh \u0111\u1ed9ng.<\/li>\n
  • Chatbot kh\u00f3 \u0111o l\u1ee3i t\u1ee9c \u0111\u1ea7u t\u01b0 (ROI) theo k\u1ebft qu\u1ea3 \u0111\u1ea7u ra (outcome):<\/strong> b\u1ea1n \u0111o \u0111\u01b0\u1ee3c s\u1ed1 l\u01b0\u1ee3t h\u1ecfi \u0111\u00e1p, nh\u01b0ng kh\u00f3 \u0111o \u0111\u01b0\u1ee3c phi\u1ebfu x\u1eed l\u00fd (ticket) \u0111\u00f3ng nhanh h\u01a1n bao nhi\u00eau hay ki\u1ec3m th\u1eed h\u1ed3i quy (regression) gi\u1ea3m bao nhi\u00eau.<\/li>\n
  • Chatbot d\u1ec5 tr\u1edf th\u00e0nh \u201cl\u1edbp giao di\u1ec7n m\u1edbi\u201d ch\u1ee9 ch\u01b0a ph\u1ea3i \u201cl\u1edbp v\u1eadn h\u00e0nh m\u1edbi\u201d:<\/strong> n\u1ebfu quy tr\u00ecnh ph\u00eda sau kh\u00f4ng thay \u0111\u1ed5i, AI ch\u1ec9 l\u00e0m giao di\u1ec7n \u0111\u1eb9p h\u01a1n.<\/li>\n<\/ul>\n

    Ch\u00ednh v\u00ec v\u1eady, nhi\u1ec1u doanh nghi\u1ec7p \u0111\u00e3 nh\u1eadn ra r\u1eb1ng b\u00e0i to\u00e1n ti\u1ebfp theo kh\u00f4ng ph\u1ea3i tri\u1ec3n khai th\u00eam chatbot m\u1edbi. B\u00e0i to\u00e1n l\u00e0 ch\u1ecdn \u0111\u00fang quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow) c\u00f3 th\u1ec3 \u0111\u01b0\u1ee3c \u201cagent h\u00f3a\u201d m\u1ed9t ph\u1ea7n, r\u1ed3i thi\u1ebft k\u1ebf quy\u1ec1n h\u1ea1n, h\u00e0ng r\u00e0o ki\u1ec3m so\u00e1t (guardrail) v\u00e0 c\u00e1ch \u0111o l\u01b0\u1eddng ph\u00f9 h\u1ee3p.<\/p>\n

    \"\"<\/p>\n

     <\/p>\n

    <\/p>\n

    AI agent thay \u0111\u1ed5i b\u00e0i to\u00e1n b\u1eb1ng c\u00e1ch n\u00e0o trong v\u1eadn h\u00e0nh doanh nghi\u1ec7p?<\/h2>\n

    \u0110i\u1ec3m m\u1ea1nh c\u1ee7a agent kh\u00f4ng ph\u1ea3i l\u00e0 m\u00f4 h\u00ecnh m\u1ea1nh h\u01a1n chatbot m\u1ed9t ch\u00fat. \u0110i\u1ec3m m\u1ea1nh n\u1eb1m \u1edf c\u00e1ch t\u1ed5 ch\u1ee9c c\u00f4ng vi\u1ec7c<\/strong>. M\u1ed9t agent t\u1ed1t th\u01b0\u1eddng ho\u1ea1t \u0111\u1ed9ng theo b\u1ed1n l\u1edbp:<\/p>\n

      \n
    1. Nh\u1eadn m\u1ee5c ti\u00eau r\u00f5 r\u00e0ng:<\/strong> v\u00ed d\u1ee5 \u201cx\u1eed l\u00fd phi\u1ebfu x\u1eed l\u00fd (ticket) ho\u00e0n ti\u1ec1n \u0111\u00fang th\u1ecfa thu\u1eadn m\u1ee9c d\u1ecbch v\u1ee5 (SLA)\u201d hay \u201c\u0111\u00e1nh gi\u00e1 ph\u1ea1m vi ki\u1ec3m th\u1eed h\u1ed3i quy (regression) c\u1ee7a b\u1ea3n d\u1ef1ng m\u1edbi (build)\u201d.<\/li>\n
    2. L\u1eadp k\u1ebf ho\u1ea1ch \u0111a b\u01b0\u1edbc:<\/strong> x\u00e1c \u0111\u1ecbnh c\u1ea7n l\u1ea5y d\u1eef li\u1ec7u \u1edf \u0111\u00e2u, g\u1ecdi c\u00f4ng c\u1ee5 n\u00e0o, ph\u1ee5 thu\u1ed9c g\u00ec, khi n\u00e0o ph\u1ea3i ch\u1edd ph\u00ea duy\u1ec7t.<\/li>\n
    3. Th\u1ef1c thi trong ranh gi\u1edbi ch\u00ednh s\u00e1ch (policy):<\/strong> ch\u1ec9 l\u00e0m nh\u1eefng h\u00e0nh \u0111\u1ed9ng \u0111\u00e3 \u0111\u01b0\u1ee3c cho ph\u00e9p, l\u01b0u l\u1ea1i nh\u1eadt k\u00fd (log), tr\u1ea1ng th\u00e1i v\u00e0 quy\u1ebft \u0111\u1ecbnh.<\/li>\n
    4. B\u00e0n giao ngo\u1ea1i l\u1ec7 cho con ng\u01b0\u1eddi:<\/strong> khi \u0111\u1ed9 tin c\u1eady (confidence) th\u1ea5p, d\u1eef li\u1ec7u thi\u1ebfu, ho\u1eb7c t\u00e1c \u0111\u1ed9ng kinh doanh cao, agent ph\u1ea3i d\u1eebng v\u00e0 chuy\u1ec3n cho ng\u01b0\u1eddi ph\u1ee5 tr\u00e1ch.<\/li>\n<\/ol>\n

      C\u00e1ch ti\u1ebfp c\u1eadn n\u00e0y \u0111\u1eb7c bi\u1ec7t ph\u00f9 h\u1ee3p v\u1edbi nh\u1eefng quy tr\u00ecnh \u201cnhi\u1ec1u \u0111i\u1ec3m ph\u00ea duy\u1ec7t (approval-heavy)\u201d v\u00e0 \u201cnh\u1ea1y v\u1ec1 th\u1eddi gian (timing-sensitive)\u201d nh\u01b0 Gartner m\u00f4 t\u1ea3 trong th\u00e1ng 4\/2026<\/strong>: c\u00e1c b\u01b0\u1edbc l\u1eb7p l\u1ea1i nhi\u1ec1u, ph\u1ee5 thu\u1ed9c d\u1eef li\u1ec7u t\u1eeb nhi\u1ec1u h\u1ec7 th\u1ed1ng, c\u1ea7n ra quy\u1ebft \u0111\u1ecbnh nhanh nh\u01b0ng v\u1eabn ph\u1ea3i ki\u1ec3m so\u00e1t \u0111\u01b0\u1ee3c. (Ngu\u1ed3n: Gartner, April 2026)<\/em><\/p>\n

      Trong k\u1ef9 thu\u1eadt ph\u1ea7n m\u1ec1m (engineering) v\u00e0 \u0111\u1ea3m b\u1ea3o ch\u1ea5t l\u01b0\u1ee3ng (QA), \u0111\u00e2y l\u00e0 khu v\u1ef1c r\u1ea5t \u0111\u00e1ng ch\u00fa \u00fd. Agent kh\u00f4ng thay tester hay QA lead. Nh\u01b0ng agent c\u00f3 th\u1ec3 nh\u1eadn ph\u1ea7n vi\u1ec7c t\u1ed1n th\u1eddi gian m\u00e0 \u0111\u1ed9 ph\u00e1n \u0111o\u00e1n \u1edf m\u1ee9c trung b\u00ecnh: ph\u00e2n lo\u1ea1i b\u00e1o c\u00e1o l\u1ed7i (bug report), t\u1ea1o b\u1ea3n nh\u00e1p ki\u1ec3m th\u1eed (test draft) t\u1eeb y\u00eau c\u1ea7u (requirement), g\u1ee3i \u00fd v\u00f9ng r\u1ee7i ro (risk area) sau thay \u0111\u1ed5i m\u00e3 ngu\u1ed3n (code change), t\u1ed5ng h\u1ee3p b\u1eb1ng ch\u1ee9ng ki\u1ec3m th\u1eed (test evidence), ho\u1eb7c r\u00e0 so\u00e1t xem b\u1ea3n d\u1ef1ng n\u00e0o (build) c\u1ea7n con ng\u01b0\u1eddi xem l\u1ea1i (human review) tr\u01b0\u1edbc khi ph\u00e1t h\u00e0nh (release).<\/p>\n

       <\/p>\n

      \u0110i\u1ec1u n\u00e0y t\u00e1c \u0111\u1ed9ng g\u00ec \u0111\u1ebfn CTO, QA Lead v\u00e0 engineering team?<\/h2>\n

      \u1ea2nh h\u01b0\u1edfng l\u1edbn nh\u1ea5t l\u00e0 c\u00e1ch b\u1ea1n thi\u1ebft k\u1ebf \u0111\u1ed9i ng\u0169 v\u00e0 ch\u1ec9 s\u1ed1 hi\u1ec7u qu\u1ea3 ch\u00ednh (KPI). Tr\u01b0\u1edbc \u0111\u00e2y, AI th\u01b0\u1eddng \u0111\u01b0\u1ee3c mua \u0111\u1ec3 gi\u00fap t\u1eebng ng\u01b0\u1eddi l\u00e0m nhanh h\u01a1n. V\u1edbi agent, tr\u1ecdng t\u00e2m chuy\u1ec3n sang c\u00e2u h\u1ecfi: quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow) n\u00e0o c\u00f3 th\u1ec3 giao cho ph\u1ea7n m\u1ec1m ch\u1ea1y tr\u01b0\u1edbc, v\u00e0 con ng\u01b0\u1eddi n\u00ean gi\u1eef l\u1ea1i ph\u1ea7n quy\u1ebft \u0111\u1ecbnh n\u00e0o.<\/p>\n

      V\u1edbi CTO, l\u1ee3i \u00edch n\u1eb1m \u1edf ch\u1ed7 gi\u1ea3m \u0111\u1ed9 tr\u1ec5 ra quy\u1ebft \u0111\u1ecbnh (decision latency) gi\u1eefa c\u00e1c h\u1ec7 th\u1ed1ng v\u1ed1n \u0111ang r\u1eddi r\u1ea1c. V\u1edbi Engineering Manager, l\u1ee3i \u00edch n\u1eb1m \u1edf vi\u1ec7c c\u1eaft nh\u1eefng b\u01b0\u1edbc trung gian khi\u1ebfn \u0111\u1ed9i ng\u0169 (team) b\u1ecb k\u1eb9t gi\u1eefa h\u1ecdp, phi\u1ebfu x\u1eed l\u00fd (ticket) v\u00e0 b\u00e0n giao (handoff). V\u1edbi QA Lead, l\u1ee3i \u00edch r\u00f5 nh\u1ea5t l\u00e0 \u0111\u01b0a ki\u1ec3m th\u1eed ti\u1ebfn g\u1ea7n h\u01a1n v\u1edbi ng\u1eef c\u1ea3nh thay \u0111\u1ed5i th\u1ef1c t\u1ebf c\u1ee7a s\u1ea3n ph\u1ea9m, thay v\u00ec ch\u1ea1y m\u1ed9t b\u1ed9 ki\u1ec3m th\u1eed (test) c\u1ed1 \u0111\u1ecbnh r\u1ed3i ch\u1edd ng\u01b0\u1eddi t\u1ed5ng h\u1ee3p.<\/p>\n

      Nh\u01b0ng \u0111\u1ed5i l\u1ea1i, team c\u0169ng ph\u1ea3i n\u00e2ng chu\u1ea9n \u1edf ba \u0111i\u1ec3m:<\/p>\n

        \n
      • Kh\u1ea3 n\u0103ng quan s\u00e1t (observability):<\/strong> ph\u1ea3i nh\u00ecn \u0111\u01b0\u1ee3c agent \u0111\u00e3 l\u00e0m g\u00ec, d\u1ef1a tr\u00ean d\u1eef li\u1ec7u n\u00e0o, v\u00e0 t\u1ea1i sao n\u00f3 ch\u1ecdn h\u01b0\u1edbng \u0111\u00f3.<\/li>\n
      • Qu\u1ea3n tr\u1ecb (governance):<\/strong> ph\u1ea3i \u0111\u1ecbnh ngh\u0129a r\u00f5 quy\u1ec1n c\u1ee7a agent, d\u1eef li\u1ec7u n\u00e0o \u0111\u01b0\u1ee3c ch\u1ea1m v\u00e0o, b\u01b0\u1edbc n\u00e0o b\u1eaft bu\u1ed9c c\u00f3 ph\u00ea duy\u1ec7t c\u1ee7a con ng\u01b0\u1eddi (human approval).<\/li>\n
      • \u0110\u00e1nh gi\u00e1 (evaluation):<\/strong> kh\u00f4ng th\u1ec3 \u0111\u00e1nh gi\u00e1 agent ch\u1ec9 b\u1eb1ng c\u1ea3m gi\u00e1c \u201ctr\u1ea3 l\u1eddi kh\u00e1 \u1ed5n\u201d; ph\u1ea3i \u0111o theo th\u1ecfa thu\u1eadn m\u1ee9c d\u1ecbch v\u1ee5 (SLA), t\u1ef7 l\u1ec7 l\u1ed7i l\u1ecdt (defect leakage), th\u1eddi gian ho\u00e0n th\u00e0nh chu k\u1ef3 (cycle time), t\u1ef7 l\u1ec7 d\u01b0\u01a1ng t\u00ednh gi\u1ea3 (false positive) v\u00e0 chi ph\u00ed v\u1eadn h\u00e0nh.<\/li>\n<\/ul>\n

        KPMG cho bi\u1ebft 57% l\u00e3nh \u0111\u1ea1o<\/strong> hi\u1ec7n k\u1ef3 v\u1ecdng con ng\u01b0\u1eddi s\u1ebd qu\u1ea3n l\u00fd v\u00e0 \u0111i\u1ec1u ph\u1ed1i AI agent<\/strong>, kh\u00f4ng ph\u1ea3i ng\u01b0\u1ee3c l\u1ea1i. (Ngu\u1ed3n: KPMG, Q1 2026)<\/em> \u0110\u00e2y l\u00e0 d\u1ea5u hi\u1ec7u cho th\u1ea5y vai tr\u00f2 c\u1ee7a team k\u1ef9 thu\u1eadt s\u1ebd d\u1ecbch chuy\u1ec3n t\u1eeb \u201cng\u01b0\u1eddi tr\u1ef1c ti\u1ebfp l\u00e0m h\u1ebft\u201d sang \u201cng\u01b0\u1eddi thi\u1ebft k\u1ebf, gi\u00e1m s\u00e1t v\u00e0 ch\u1ec9nh h\u00e0nh vi h\u1ec7 th\u1ed1ng\u201d.<\/p>\n

        \"\"<\/p>\n

         <\/p>\n

        <\/p>\n

        N\u1ebfu doanh nghi\u1ec7p c\u1ee7a b\u1ea1n mu\u1ed1n chuy\u1ec3n t\u1eeb chatbot sang agent, n\u00ean b\u1eaft \u0111\u1ea7u th\u1ebf n\u00e0o trong 90 ng\u00e0y?<\/h2>\n

        N\u1ebfu b\u1ea1n m\u1edbi \u1edf giai \u0111o\u1ea1n chatbot n\u1ed9i b\u1ed9 ho\u1eb7c c\u00e2u h\u1ecfi th\u01b0\u1eddng g\u1eb7p (FAQ):<\/strong> h\u00e3y ch\u1ecdn m\u1ed9t quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow) nh\u1ecf nh\u01b0ng c\u00f3 \u0111\u1ea7u ra \u0111o \u0111\u01b0\u1ee3c, v\u00ed d\u1ee5 ph\u00e2n lo\u1ea1i phi\u1ebfu x\u1eed l\u00fd (ticket), t\u1ed5ng h\u1ee3p bi\u00ean b\u1ea3n h\u1ecdp, ho\u1eb7c t\u1ea1o b\u1ea3n nh\u00e1p ki\u1ec3m th\u1eed (test draft) t\u1eeb y\u00eau c\u1ea7u thay \u0111\u1ed5i (change request). M\u1ee5c ti\u00eau l\u00e0 \u0111o \u0111\u01b0\u1ee3c th\u1eddi gian ho\u00e0n th\u00e0nh chu k\u1ef3 (cycle time), t\u1ec9 l\u1ec7 c\u1ea7n ng\u01b0\u1eddi s\u1eeda, v\u00e0 t\u00e1c \u0111\u1ed9ng l\u00ean th\u1ecfa thu\u1eadn m\u1ee9c d\u1ecbch v\u1ee5 (SLA).<\/p>\n

        N\u1ebfu b\u1ea1n \u0111\u00e3 c\u00f3 tr\u1ee3 l\u00fd \u0111\u1ed3ng h\u00e0nh (copilot) ho\u1eb7c chatbot t\u00edch h\u1ee3p d\u1eef li\u1ec7u n\u1ed9i b\u1ed9:<\/strong> b\u01b0\u1edbc ti\u1ebfp theo kh\u00f4ng ph\u1ea3i l\u00e0 m\u1edf th\u00eam tr\u01b0\u1eddng h\u1ee3p s\u1eed d\u1ee5ng (use case) d\u1ea1ng chat. H\u00e3y th\u00eam m\u1ed9t l\u1edbp \u0111i\u1ec1u ph\u1ed1i (orchestration) c\u00f3 ch\u00ednh s\u00e1ch (policy) r\u00f5 r\u00e0ng: agent \u0111\u01b0\u1ee3c ph\u00e9p \u0111\u1ecdc g\u00ec, ghi g\u00ec, khi n\u00e0o ph\u1ea3i xin ph\u00ea duy\u1ec7t (approval), nh\u1eadt k\u00fd (log) l\u01b0u \u1edf \u0111\u00e2u, ai ch\u1ecbu tr\u00e1ch nhi\u1ec7m khi \u0111\u1ea7u ra (output) sai.<\/p>\n

        N\u1ebfu b\u1ea1n \u0111\u00e3 th\u1eed agent \u1edf m\u1ed9t v\u00e0i quy tr\u00ecnh:<\/strong> \u0111\u1eebng m\u1edf r\u1ed9ng quy m\u00f4 (scale) tr\u01b0\u1edbc khi c\u00f3 b\u1ed9 ch\u1ec9 s\u1ed1 t\u1ed1i thi\u1ec3u. \u00cdt nh\u1ea5t c\u1ea7n c\u00f3: th\u1eddi gian x\u1eed l\u00fd tr\u01b0\u1edbc v\u00e0 sau agent, t\u1ec9 l\u1ec7 ngo\u1ea1i l\u1ec7 (exception), chi ph\u00ed m\u1ed7i quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow), m\u1ee9c \u0111\u1ed9 can thi\u1ec7p th\u1ee7 c\u00f4ng, v\u00e0 c\u00e1c s\u1ef1 c\u1ed1 (incident) li\u00ean quan t\u1edbi d\u1eef li\u1ec7u ho\u1eb7c quy\u1ec1n truy c\u1eadp.<\/p>\n

        Th\u1ee9 t\u1ef1 \u01b0u ti\u00ean h\u1ee3p l\u00fd th\u01b0\u1eddng l\u00e0:<\/p>\n

          \n
        1. Ch\u1ecdn quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow) l\u1eb7p l\u1ea1i, nhi\u1ec1u \u0111i\u1ec3m b\u00e0n giao (handoff), \u00edt r\u1ee7i ro ph\u00e1p l\u00fd.<\/strong><\/li>\n
        2. Thi\u1ebft k\u1ebf h\u00e0ng r\u00e0o ki\u1ec3m so\u00e1t (guardrail) tr\u01b0\u1edbc khi vi\u1ebft c\u00e2u l\u1ec7nh (prompt).<\/strong><\/li>\n
        3. \u0110o k\u1ebft qu\u1ea3 \u0111\u1ea7u ra kinh doanh (business outcome) tr\u01b0\u1edbc khi m\u1edf r\u1ed9ng ph\u1ea1m vi.<\/strong><\/li>\n
        4. N\u00e2ng k\u1ef9 n\u0103ng (upskill) cho ng\u01b0\u1eddi qu\u1ea3n l\u00fd quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow), kh\u00f4ng ch\u1ec9 \u0111\u00e0o t\u1ea1o (training) ng\u01b0\u1eddi d\u00f9ng cu\u1ed1i.<\/strong><\/li>\n<\/ol>\n

          Theo Microsoft, 78% l\u00e3nh \u0111\u1ea1o<\/strong> \u0111ang c\u00e2n nh\u1eafc tuy\u1ec3n c\u00e1c vai tr\u00f2 chuy\u00ean bi\u1ec7t cho AI \u0111\u1ec3 chu\u1ea9n b\u1ecb cho giai \u0111o\u1ea1n t\u1edbi. (Ngu\u1ed3n: Microsoft Work Trend Index 2025)<\/em> Nh\u01b0ng v\u1edbi ph\u1ea7n l\u1edbn doanh nghi\u1ec7p, b\u01b0\u1edbc \u0111i \u0111\u00fang ngay l\u00fac n\u00e0y ch\u01b0a ph\u1ea3i tuy\u1ec3n m\u1ed9t \u0111\u1ed9i \u201ck\u1ef9 s\u01b0 agent (agent engineers)\u201d ri\u00eang. B\u01b0\u1edbc \u0111\u00fang l\u00e0 x\u00e1c \u0111\u1ecbnh \u0111\u1ed9i n\u00e0o s\u1ebd l\u00e0m ch\u1ee7 ch\u00ednh s\u00e1ch (policy), d\u1eef li\u1ec7u v\u00e0 ch\u1ec9 s\u1ed1 \u0111o l\u01b0\u1eddng (metric) c\u1ee7a t\u1eebng quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow) \u0111\u01b0\u1ee3c agent h\u00f3a.<\/p>\n

           <\/p>\n

          K\u1ebft lu\u1eadn: chuy\u1ec3n \u0111\u1ed5i ngay kh\u00f4ng c\u00f3 ngh\u0129a l\u00e0 thay m\u1ecdi chatbot b\u1eb1ng agent trong m\u1ed9t qu\u00fd<\/h2>\n

          Nh\u1eadn \u0111\u1ecbnh c\u1ee7a ch\u00fang t\u00f4i l\u00e0: doanh nghi\u1ec7p c\u1ea7n chuy\u1ec3n \u0111\u1ed5i ngay kh\u00f4ng ph\u1ea3i v\u00ec chatbot \u0111\u00e3 v\u00f4 d\u1ee5ng, m\u00e0 v\u00ec chatbot ch\u1ec9 c\u00f2n l\u00e0 l\u1edbp \u0111\u1ea7u c\u1ee7a h\u00e0nh tr\u00ecnh AI<\/strong>. Gi\u00e1 tr\u1ecb l\u1edbn h\u01a1n \u0111ang d\u1ecbch chuy\u1ec3n sang c\u00e1c h\u1ec7 th\u1ed1ng c\u00f3 th\u1ec3 th\u1ef1c thi c\u00f4ng vi\u1ec7c, ph\u1ed1i h\u1ee3p nhi\u1ec1u b\u01b0\u1edbc, v\u00e0 ch\u1ecbu r\u00e0ng bu\u1ed9c ki\u1ec3m so\u00e1t nh\u01b0 m\u1ed9t th\u00e0nh ph\u1ea7n th\u1eadt c\u1ee7a v\u1eadn h\u00e0nh doanh nghi\u1ec7p.<\/p>\n

          \u201cNgay\u201d \u1edf \u0111\u00e2y n\u00ean \u0111\u01b0\u1ee3c hi\u1ec3u l\u00e0 b\u1eaft \u0111\u1ea7u thi\u1ebft k\u1ebf l\u1ea1i workflow v\u00e0 governance t\u1eeb b\u00e2y gi\u1edd<\/strong>, kh\u00f4ng ph\u1ea3i v\u1ed9i v\u00e0ng t\u1ef1 \u0111\u1ed9ng h\u00f3a to\u00e0n b\u1ed9. Nh\u1eefng \u0111\u1ed9i \u0111i tr\u01b0\u1edbc s\u1ebd kh\u00f4ng th\u1eafng v\u00ec h\u1ecd c\u00f3 nhi\u1ec1u chatbot h\u01a1n. H\u1ecd th\u1eafng v\u00ec h\u1ecd ch\u1ecdn \u0111\u00fang quy tr\u00ecnh \u0111\u1ec3 giao cho agent, \u0111o \u0111\u00fang k\u1ebft qu\u1ea3, v\u00e0 gi\u1eef con ng\u01b0\u1eddi \u1edf \u0111\u00fang \u0111i\u1ec3m ki\u1ec3m so\u00e1t.<\/p>\n

           <\/p>\n

          Nh\u00ecn chung l\u1ea1i:<\/strong><\/p>\n

            \n
          • Th\u1ecb tr\u01b0\u1eddng \u0111ang d\u1ecbch chuy\u1ec3n kh\u1ecfi AI h\u1ed7 tr\u1ee3 \u0111\u01a1n thu\u1ea7n sang AI cam k\u1ebft k\u1ebft qu\u1ea3 quy tr\u00ecnh c\u00f4ng vi\u1ec7c (workflow); Gartner d\u1ef1 b\u00e1o h\u01a1n m\u1ed9t n\u1eeda doanh nghi\u1ec7p s\u1ebd \u01b0u ti\u00ean h\u01b0\u1edbng n\u00e0y v\u00e0o n\u0103m 2028.<\/li>\n
          • Chatbot v\u1eabn h\u1eefu \u00edch, nh\u01b0ng tr\u1ea7n ROI xu\u1ea5t hi\u1ec7n khi doanh nghi\u1ec7p c\u1ea7n AI \u0111i qua nhi\u1ec1u b\u01b0\u1edbc, nhi\u1ec1u h\u1ec7 th\u1ed1ng v\u00e0 nhi\u1ec1u l\u1edbp ph\u00ea duy\u1ec7t.<\/li>\n
          • \u0110\u1ec3 chuy\u1ec3n t\u1eeb chatbot sang agent hi\u1ec7u qu\u1ea3, doanh nghi\u1ec7p c\u1ea7n \u0111\u1ed3ng th\u1eddi n\u00e2ng chu\u1ea9n \u1edf workflow design, governance v\u00e0 evaluation.<\/li>\n<\/ul>\n

            Ngu\u1ed3n tham kh\u1ea3o:<\/strong><\/p>\n

              \n
            • Gartner Newsroom, Gartner Predicts 40% of Enterprise Apps Will Feature Task-Specific AI Agents by 2026, Up from Less Than 5% in 2025<\/em>, 26\/08\/2025.<\/li>\n
            • Gartner Newsroom, Gartner Expects Most Enterprises to Abandon Assistive AI for Outcome-Focused Workflow by 2028<\/em>, 02\/04\/2026.<\/li>\n
            • Microsoft WorkLab, 2025: The year the Frontier Firm is born<\/em>, 23\/04\/2025.<\/li>\n
            • KPMG, AI Quarterly Pulse Survey – Q1 2026<\/em>.<\/li>\n
            • Deloitte Southeast Asia, Agentic and physical AI set for rapid growth in Singapore in the next two years<\/em>, 03\/02\/2026.<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"

              AI agent doanh nghi\u1ec7p \u0111ang v\u01b0\u1ee3t kh\u1ecfi vai tr\u00f2 chatbot h\u1ed7 tr\u1ee3 \u0111\u1ec3 \u0111i v\u00e0o workflow th\u1ef1c thi. \u0110\u00e2y l\u00e0 l\u00fac CTO v\u00e0 QA Lead c\u1ea7n thi\u1ebft k\u1ebf l\u1ea1i quy tr\u00ecnh.<\/p>\n","protected":false},"author":3,"featured_media":1619,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[25,23],"tags":[55,46,82,83,85,60,81,84],"class_list":["post-1613","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tin-cong-nghe","category-tin-tuc","tag-agenticai","tag-ai-agent","tag-chatbot","tag-doanhngiep","tag-qaengineering","tag-taas","tag-tincongnghe","tag-workflowautomation"],"acf":[],"_links":{"self":[{"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/posts\/1613","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/comments?post=1613"}],"version-history":[{"count":12,"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/posts\/1613\/revisions"}],"predecessor-version":[{"id":1666,"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/posts\/1613\/revisions\/1666"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/media\/1619"}],"wp:attachment":[{"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/media?parent=1613"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/categories?post=1613"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/bbotech.vn\/vi\/wp-json\/wp\/v2\/tags?post=1613"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}