Forward Deployment Engineer

NadaLab; Malaysia; On-site; MYR 120,000 - 240,000; Full time

Job summary

NadaLab in Malaysia is seeking a hands-on FDE to own deployment from POC to production within SEA, embedded with clients to bridge our platform with real-world telephony, data and workflows.

You will diagnose live calls, tune persona and language (BM, Manglish, Singlish, Mandarin, Cantonese) and integrate APIs, telephony, CRM and data feeds while delivering rapid AI-assisted configurations.

Qualifications

• Own a deployment as the primary technical person through to its agreed success metric.
• Hear what is wrong in a live voice interaction and improve the behaviour—not only the code.
• Tune persona, register, tone, language and code-switching for a specific client and use case.
• Connect APIs, telephony, CRM, data and messaging systems with strong integration instincts.
• Use AI tooling and agentic workflows to move rapidly from brief to working solution.

Responsibilities

• End-to-end deployment ownership
• Real-time Voice AI tuning and diagnosis
• Persona engineering and localisation
• Systems integration
• Customer-facing technical translation
• AI-leveraged build speed
• Field-to-platform feedback

Job description

Location

Malaysia / SEA-wide

Working model

Client-embedded

Mission

The role

Own the solution until it reaches the agreed success metric.

The FDE is the linchpin of NadaLab’s POC-to-paid motion. You embed with clients and bridge our platform with the reality of customer telephony, data, language and workflows.

This is a hands-on, customer-facing engineering role for someone who can operate independently in ambiguity.

Key responsibilities

End-to-end deployment ownership

Take signed POCs through configuration, integration, tuning, UAT, go-live and early-life support.

Real-time Voice AI tuning and diagnosis

Listen to live calls and transcripts to diagnose latency, false barge-in, interruption, containment and persona issues.

Persona engineering and localisation

Engineer voice, register, tone, language and code-switching for the client’s use case, including BM, Manglish, Singlish, Mandarin and Cantonese.

Systems integration

Connect telephony, CRM, data feeds and messaging channels by reading documentation and making systems work together.

Customer-facing technical translation

Turn ambiguous operational needs into working configurations while managing expectations and building trust.

AI-leveraged build speed

Use AI tooling and agentic workflows to configure, test and ship quickly while prioritising production outcomes.

Field-to-platform feedback

Turn recurring field problems into clear, prioritised product and platform requirements.

What we are looking for

Practical engineering judgement with client credibility.

Own a deployment as the primary technical person through to its agreed success metric.

Hear what is wrong in a live voice interaction and improve the behaviour—not only the code.

Tune persona, register, tone, language and code-switching for a specific client and use case.

Connect APIs, telephony, CRM, data and messaging systems with strong integration instincts.

Work credibly with operations and CX teams and translate vague requirements into configuration.

Use AI tooling and agentic workflows to move rapidly from brief to working solution.

Convert field insight into actionable product feedback and platform priorities.

Technical environment

The role sits at the intersection of Voice AI, LLMs, real-time communications and enterprise integration. We value practical engineering ability over a particular technology stack.

Python or scripting REST APIs Webhooks JSON Telephony / SIP STT / TTS LLM APIs AI agent frameworks CRM integrations WhatsApp Business API Cloud services Git Production debugging

How we assess

Show us how you think in the work itself.

The process focuses on diagnosis, localisation, deployment judgement and customer translation rather than abstract interview puzzles.

01 Voice diagnosis

Diagnose a call that went wrong across latency, interruption handling, containment and persona behaviour.

02 Localisation

Explain how you would make a Manglish agent sound less robotic across persona, prompting, speech and conversation design.

03 Live deployment exercise

Scope an integration and first persona configuration from a client brief, showing ownership, speed and technical judgement.

What success looks like

A stable operating solution—and a stronger platform.

You take ownership from POC to production, achieve agreed business and conversation-quality metrics, build client confidence and leave behind a stable solution. You also make NadaLab better by feeding high-quality field insight into product and R&D.