DXTech

Challenge Story
A fast-growing tech company in Vietnam with a customer base spanning Japan, Korea, and Southeast Asia faced growing pains in managing customer conversations:

• Sales Team: Struggled to balance active listening with manual note-taking during calls. Important client details, objections, or follow-ups were sometimes lost or incomplete.

• Customer Support: Repeated issues arose because past customer context wasn’t consistently recorded or shared between agents.

• Team Leads: Needed visibility into deal progress and customer satisfaction trends, but post-call summaries were inconsistent and took hours to compile.

With 50+ customer meetings per week across multiple time zones, the team urgently needed a scalable solution to document conversations and extract insights without manual effort.

DX Tech Solution
Dxtech built Notetaker AI, a smart meeting assistant to automate post-call documentation and provide structured insights:

• Call Integration: The AI joined scheduled Zoom, Google Meet, or Microsoft Teams calls as a silent participant.

• Multilingual Transcription: Accurately transcribed conversations in Vietnamese, English, Japanese, and Korean using AI speech models.

• Conversation Summarization: Used large language models (LLMs) to extract key points, client concerns, commitments made, and action items.

• CRM Sync: Notes were auto-tagged by customer name and synced directly into the company’s CRM and internal chat (e.g., Slack, LINE).

• Customization: Teams could define summary formats (e.g., STAR method, Issue–Solution–Outcome) and keywords to track (e.g., “renewal,” “budget,” “problem”).

Results

• Reduced manual note-taking workload by 90% across the sales and support teams.

• Increased data consistency, with over 95% of calls now documented in the CRM.

• Cut post-call processing time from 20–30 minutes to under 5 minutes.

• Improved customer satisfaction by ensuring every follow-up was timely and based on accurate context.

• Enables sales managers to review conversation summaries and deal health across 200+ clients/month.

Challenge Story
A Japanese manufacturing company with operations in Vietnam faced serious communication gaps during cross-border meetings:

• Japanese Executives: Preferred speaking in Japanese, especially for complex technical and financial topics. They found English insufficient and Vietnamese completely inaccessible.

• Vietnamese Team: Struggled to fully understand meeting discussions, often relying on limited English or manual note-taking by bilingual staff. This led to delays in task execution and misinterpretation of instructions.

• HR & Operations: Found it difficult to onboard new staff or scale meetings without translators. Real-time note-taking was inconsistent, and post-meeting summaries were incomplete.

Hiring human interpreters for every meeting was expensive and not scalable. Miscommunication became a growing risk as more departments joined cross-border calls.

DX Tech Solution
Dxtech developed a Meeting Translator Agent to provide real-time bilingual support for Japanese-Vietnamese meetings:

• Live Transcription: Integrated with Zoom and Google Meet, the agent captured audio streams and transcribed them live in Japanese and Vietnamese using speech-to-text models.

• Translation Layer: Transcribed content was translated in real-time using custom-tuned AI models that preserved technical terminology and business context.
  Speaker Identification: Used voice fingerprinting and meeting metadata to tag speakers (e.g., "Mr. Sato – CFO").

• Post-Meeting Summary: Automatically generated bilingual meeting minutes, highlighting key action points, decisions, and deadlines.
  Private & Secure: Complied with data privacy regulations and ensured conversations were only stored temporarily for processing.

Results

• Reduced the need for professional interpreters by 80% across weekly meetings.

• Improved comprehension and task accuracy for the Vietnamese team, resulting in a 35% reduction in follow-up clarification requests.

• Cut meeting prep and post-processing time from 2–3 hours to under 20 minutes.

• Enabled HR to onboard new team members faster with full access to multilingual meeting archives.

• Supported over 40+ meetings/month with high translation accuracy for both technical and casual dialogue.

Challenge Story
A Vietnamese state agency managing over 1 million documents annually began facing growing challenges as document categories expanded and information retrieval needs became more complex. The legacy system, optimized for basic storage and keyword search, exposed critical inefficiencies:

• Public Servants: Struggled to locate the right documents due to keyword limitations and poor metadata tagging. Document relationships were unclear, and cross-referencing between categories was nearly impossible.

• Internal IT Team: Couldn’t keep up with demands for smarter search tools. Manual tagging and categorization became unsustainable as document volume crossed 2.5 million.

• Decision-Making Units: Lost valuable time extracting key data buried deep within legal texts, reports, and policy documents — leading to slower responses and higher error rates in official work.

DX Tech Solution
Dxtech built Tanika, an AI-powered assistant that acts as an intelligent layer over existing document storage systems:

• Data Ingestion & Indexing: All files (PDFs, DOCs, scanned images) were processed with OCR and vectorized using OpenAI embeddings, then stored in a vector database.

• RAG (Retrieval-Augmented Generation): When a user asks a question, the system semantically searches the database and combines results with a language model to generate precise answers.

• Natural Language Interface: Users can ask questions like “Find the latest procurement policy update” and receive instant, context-aware answers — not just file names.

• Access Control Integration: Permissions were matched with internal user roles to ensure secure results.

Results

• Search time was reduced by 92%, from 6–10 minutes per query to under 30 seconds.

• Staff productivity increased by 30–40% in departments that adopted the tool.

• Error rates dropped by 70% when retrieving sensitive policy or compliance documents.

• The agent saved an estimated 300–400 manual hours/month across the agency.

Challenge Story
In Vietnam and across Southeast Asia, over 3 million small and medium-sized businesses (SMEs) face a silent but critical growth barrier: customer relationship management. Most SMEs operate without proper CRM tools, and over 80% store customer data in spreadsheets, notebooks, or not at all.

This leads to several pain points:

• Sales Teams: Can’t track customer history or follow up effectively. Outreach is often manual and generic, missing opportunities for personalized engagement.

• Marketing Teams: Struggle to run campaigns across platforms like Zalo, Facebook, or TikTok without automation tools. Coordination is ad hoc, and performance data is siloed.

• Business Owners: Lack a single view of customer touchpoints and miss out on upsell/cross-sell potential. They want automation but can’t afford or manage complex CRM platforms like Salesforce or HubSpot.

Without a centralized system, SMEs were losing repeat customers, overspending on ads, and failing to scale personalized outreach.

DX Tech Solution
Dxtech built Tanigo, an AI-powered CRM assistant tailored to the specific needs of small businesses in Vietnam and SEA.

• Customer Data Management: Tanigo automatically captures and stores customer data from web forms, chats, orders, and Zalo/Facebook interactions.

• Smart Segmentation: The system uses rules and AI clustering to group customers based on purchase history, interests, and engagement patterns.

• Multichannel Automation: Businesses can create campaigns that send targeted messages via Zalo, TikTok, email, or SMS — all scheduled and personalized with AI.

• Agent Interface: A lightweight, chatbot-style interface allows business owners to issue natural language commands like “Send a birthday promo to loyal customers” or “Create a drip campaign for new leads.”

• Workflow Integration: Tanigo integrates with existing eCommerce sites and POS systems via APIs or tools like Zapier/n8n, minimizing onboarding friction.

Results

• SMEs using Tanigo grew their customer re-engagement rates by 30–50% within 2–3 months

• Teams saved 20–30 hours/month on manual messaging and campaign coordination

• Response rates to campaigns improved by 3–5x, thanks to better timing and personalization

• Over 200,000 customer records have been processed and segmented across initial deployments

• Business owners reported “feeling in control” of their sales pipeline for the first time without needing a dedicated marketing hire

Head Management System Customer Service System

Overview:

Big project for big client in car production in Viet Nam.

We updated thier Head Management System  and develop Order Sheet Management – which heads will use when motobike owner go to head for motobike service like repair, replace stuffs.

• 6 months
• PM 1.0, BA 1.0, Infra 1.0, Engineer 8.0, Tester 4.0
• Language: Typescript, Javascript Framework: Node JS, React Native, React JS
• Cloud: AWS
• GitHub、Jira Software、 Slack
• Agile Development, Project Base

Project Info:

• Category: Software Development
• Client: Japan
• Industry: Business

Welfare Management System

Overview:

Welfare management system provide one–stop system for stafves and providers of client, where they can use, leverage client welfare like: Get point from good job, use point to pay foods, drinks in canteens, etc…

It also connected with ERP of client to autmation the data processing.

• 6 months
• PM 1.0, BA 1.0, Engineer 4.0, Tester 3.0
• Language: C#、Typescript
• Framework: ,NET, React JS
• Cloud: On-premise
• GitHub、Jira Software、 Slack
• Agile Development, Project Base

Project Info:

• Category: Software Development
• Client: Japan
• Industry: Business

Customer Management System (CRM) Customer Data Platform (CDP)

Overview:

We build CRM and CDP for Global Big Travel Corporation, which enablem improve customer data management of customers.

Also enhance the customer experience when use Corporation Service.

• 4 months
• PM 1.0, Designer/BA 1.0, Engineer 4.0, Tester 2.0
• Language: PHP、Typescript
• Framework: Laravel, React JS, Zalo Mini App
• Cloud: AWS
• GitHub、Jira Software、 Slack
• Agile Development, Project Base

Project Info:

• Category: Software Development
• Client: Japan
• Industry: Business

Tanigo:

Overview:
An omni-channel CRM automation marketing platform integrated with AI to enhance the experience and conversion efficiency for small and medium-sized businesses in Vietnam

Revenue: $0 - ~10 customers have committed to using and paying monthly from June 2024.
Launch: 06/2024

Project Info:

• Category: Chat Bot LLMS
• Client: Japan
• Industry: Business