About the Role

You will lead the applied ML team at VALSEA, owning the quality, direction and delivery of speech intelligence across SEA languages, accents and real-world audio conditions. This is not a research role and it is not a pure management role. You will be hands-on in production while setting the technical direction and raising the bar for the team around you.

If you thrive at the intersection of research depth and shipping velocity, and you want to own a problem that affects how machines understand how billions of people actually speak — this role is for you.

What You Will Own

• Set the technical direction for ML and speech model development across SEA languages and accents
• Lead, mentor and grow a team of ML engineers and interns
• Own the quality of speech outputs in production — accuracy, robustness, latency and cost
• Drive the team from experiments to deployed improvements without losing momentum
• Define evaluation standards, experiment frameworks and model quality benchmarks
• Collaborate with engineering to integrate models into production pipelines
• Identify systemic failure modes across languages and audio conditions and drive solutions
• Build a culture of ownership, initiative and rigorous experimentation within the team

What This Role Is Really About

You are not here to manage a research lab.

You are here to:

• Lead a team that ships measurable improvements to real production speech systems
• Raise the technical bar while keeping the team moving fast
• Make decisions under uncertainty — messy data, evolving requirements, incomplete signals
• Be the person who finds model failures before customers do
• Translate business needs into ML strategy and ML results into business impact

What We Expect From You

Leadership

• You set clear technical direction without micromanaging
• You give direct, honest feedback and make the team better for it
• You unblock engineers and remove friction from the path to shipping
• You hire and develop talent with high standards and low ego
• You represent the ML team's work clearly to founders and stakeholders

Founding Mindset

• You think in shipped improvements, not paper metrics
• You ask how will this behave in production before trying a new approach
• You act like speech quality is your personal responsibility
• You balance research depth with shipping velocity
• You do not wait for others to point out model failures — you go find them

Maturity

• You communicate clearly about what is known, unknown and risky
• You admit when an experiment failed and extract learning from it
• You stay calm under pressure when models behave unexpectedly in production
• You follow through on investigations into failure modes

What We're Looking For

• 4+ years in applied ML or speech engineering with at least 1–2 years leading a team
• Strong Python and PyTorch fundamentals
• Deep understanding of ASR, speech processing and evaluation methods
• Experience shipping ML models into real production systems
• Familiarity with GPU inference, optimisation workflows and production constraints
• Ability to define and enforce evaluation standards across a team
• Strong communicator — can translate between research, engineering and business

Strong Plus

• Experience with multilingual or low-resource speech — especially SEA languages
• Exposure to on-device or low-latency inference
• Experience building evaluation pipelines and annotation workflows
• Background in code-switching, accent robustness or noisy audio scenarios

What Success Looks Like

• Your team ships at least one measurable improvement per sprint across accuracy, robustness or latency
• You establish clear evaluation standards and experiment documentation practices
• You identify and resolve systemic failure modes across languages and audio conditions
• You develop team members who take increasing ownership over time

Who Will Thrive Here

• Technical leaders who lead by doing, not just directing
• Builders who love shipping ML to production
• Systems thinkers who see the whole pipeline, not just the model
• High-agency individuals who hold themselves accountable for real-world impact

Who Should Not Apply

• If you prefer managing without staying technically sharp
• If you avoid messy data and hard debugging
• If you prefer purely research environments detached from production
• If you need complete specifications before making decisions

What You'll Get

• Full ownership of ML direction at a speech intelligence company built for SEA
• Direct collaboration with founders and senior engineers
• A team to lead and develop from the ground up
• Meaningful equity aligned with your impact
• Remote-first flexibility

This is an equity-only role. No salary is provided at this stage.

About VALSEA

We're building the speech intelligence layer for Southeast Asia — turning real-world, accented, code-switched speech into structured, usable outputs for businesses.