ML Engineer, Voice AI
Seniority: flexible (exceptional candidates at any level)
The Role
We are looking for a talented research-minded ML engineer to join Krisp’s research team. You will become part of a small team that is highly motivated, focused on research and engineering excellence, and knows what it takes to deliver with high ownership. Everyone on the team operates with a hands-on mindset and contributes directly to outcomes, with strong prioritization and consistently high-quality execution.
The problems you’ll work on are central to Voice AI and may involve noise and background voice cancellation, accent conversion, and other voice-related challenges depending on priorities.
This role is for people who enjoy difficult problems, care about truth over ego, thrive on curiosity, and have the capacity to work independently on hard, ambiguous problems.
About the team
You will join a team of four researchers working on the frontiers of Voice AI, with backgrounds in computer science, mathematics, and physics (three PhDs: University of Geneva, University of Edinburgh, and Yerevan State University, with strong publication record in highly regarded peer-reviewed journals).
We are hands-on with coding and operate with a “do what it takes to help the team and the company to succeed” mindset. The team has access to solid infrastructure for training ML models and is supported by strong QA engineers with extensive automation, diverse test datasets, and mature methodologies – creating a fast, high-signal feedback loop.
We code in Python, use TensorFlow and PyTorch, and have a list of in-house tools and frameworks covering data handling, distributed training and computations, quality assessments, etc.
What you'll do
You will work on critical problems at the core of Voice AI and contribute directly to the mission of Krisp. The scope spans research, engineering, and productization:
- Research and prototype new methods in Voice AI (read, reason, come up with new ideas, implement, test, iterate).
- Build and train ML systems end-to-end: data – models – evaluation – iteration.
- Prepare and improve training data.
- Analyze failures rigorously (root cause, targeted experiments, clear conclusions).
- Collaborate closely with engineers and QA to optimize, validate, and ship research outputs into production.
- Improve the state of various internal tooling for experimentation, data handling, distributed computations, quality measurements, etc.
What we are looking for
We are looking for integrity, curiosity, and a collaborative spirit. Your demonstrated excellence matters far more than titles and certifications.
If what follows describes you, you’ll feel at home here:
- You are trustworthy and intellectually honest: you don’t over-claim, you’re careful with evidence, and you correct course quickly when wrong.
- You are curious and rigorous: you enjoy forming hypotheses, designing experiments, and learning deeply.
- You strive for deep understanding and craft, and you value thinking from first principles.
- You are self-driven and disciplined: you can take initiative, own ambiguous work, communicate progress, and deliver.
- You are collaborative: you share context, ask for feedback early, and find joy in helping your teammates succeed.
- You are hands-on: you write code that others can rely on, and you care about clarity, maintainability, and reproducibility.
- You are fast and resourceful: when blocked, you find pragmatic ways forward.
- Under deadlines and setbacks, you stay calm and resolute; you’re willing to push hard when it matters.
Technical background
- Strong foundations in at least one of: math, CS, physics (or related).
- Solid coding skills (typically Python, or C, C++, or similar).
- Ability to explain technical details clearly and concisely.
- Background in modern deep learning is a plus.
Not every exceptional candidate has a standard ML background. If you have a strong record in a technically demanding discipline, we want to hear from you: we have seen over and over again that deep competence transfers well.
