Compensation: TBD, based on education, experience, role fit, startup needs, and interview process.
Job summary
As a Data Scientist, you will support one or more startups connected to Metasoft Capital by helping founders turn early ambition into practical execution. This role is built for remote startup environments where priorities can change quickly, information may be incomplete, and useful work depends on clear writing, steady ownership, and good judgment.
The role focuses on the core responsibilities described for this role. You may work with founders, product teams, technical teams, operations, customers, partners, or investors depending on the startup assignment. The goal is to help the company build, deploy, evaluate, secure, govern, and scale AI products, models, agents, platforms, and enterprise AI transformation programs, while building habits that make the team more reliable as it grows.
Core responsibilities
- Own day-to-day work connected to this function, including work that uses statistics and machine learning to extract insights, run experiments, and build predictive models, keeping priorities clear and progress visible.
- Work closely with founders, product, engineering, operations, customer-facing teams, partners, or advisors depending on the startup's needs.
- Create clear notes, requirements, decisions, checklists, documentation, or reports so work does not depend on memory or guesswork.
- Identify risks, dependencies, tradeoffs, and practical next steps before small problems become expensive startup mistakes.
- Help the startup move from ideas and experiments toward usable systems, measurable traction, and durable operating habits.
- Connect AI work to real users, business outcomes, safety expectations, evaluation criteria, and measurable product value.
- Design AI systems that can be observed, evaluated, improved, secured, and governed after launch, not only demonstrated once.
- Work across model selection, prompts, retrieval, data pipelines, APIs, agents, infrastructure, UX, policies, and adoption workflows as needed.
- Identify risks such as hallucination, bias, data leakage, model drift, adversarial behavior, cost growth, and unclear accountability.
- Turn messy information into clear analysis, written findings, dashboards, recommendations, or experiments that help the startup make better decisions.
- Question assumptions, check data quality, and explain uncertainty honestly instead of presenting weak evidence as certainty.
Education and experience requirements
- Education or experience in AI, machine learning, data science, computer science, statistics, software engineering, research, or equivalent applied AI work is helpful.
- Strong projects, research, production deployments, model evaluations, open-source contributions, or enterprise AI implementation experience may be considered.
- Experience with LLMs, RAG, embeddings, agents, evaluation frameworks, ML pipelines, model APIs, MLOps, GPUs, or AI product delivery is valuable.
- Ability to separate useful AI from hype and explain model capabilities, limitations, costs, and risks in plain language.
- Interest in responsible AI, user trust, security, fairness, privacy, and safe deployment practices is important.
- Candidates should be ready to share examples of relevant work, projects, portfolios, case studies, writing samples, code samples, dashboards, campaigns, processes, or other evidence of ability when applicable.
Helpful skills and tools
- Python, ML frameworks, LLM APIs, vector databases, RAG systems, evaluation tools, MLOps platforms, GPUs, notebooks, orchestration frameworks, and monitoring tools.
- Relevant hands-on experience, training, or strong practical interest in ai industry & applied intelligence work and startup environments.
- Ability to communicate clearly with technical, non-technical, clinical, creative, commercial, or executive teammates as needed.
- Strong ownership mindset, reliability in remote work, and comfort working with incomplete information.
- Good judgment around quality, speed, cost, security, privacy, compliance, ethics, and customer impact.
- Curiosity, humility, and willingness to keep learning as the startup changes direction, grows, or enters new markets.
- Clear written communication, organized files, practical documentation, and reliable follow-through are important because most teams operate remotely.
Remote startup working style
- Work asynchronously when possible, document decisions, and keep teammates updated on progress, blockers, risks, and next steps.
- Be comfortable with early-stage ambiguity, changing priorities, limited resources, and the need to learn quickly from customers and teammates.
- Use sound judgment to decide when to move fast, when to ask for help, when to slow down, and when to protect quality or trust.
- Respect confidentiality, founder context, customer information, and the trust required to support early-stage companies.
Success in this role looks like
- Successful work produces useful AI features, reliable models, stronger model evaluation, safer deployments, better data pipelines, faster customer adoption, and measurable productivity gains.
- AI features produce reliable value and can be measured against clear standards.
- Model risks, costs, and quality are tracked instead of ignored.
- The startup builds AI systems users can trust and teams can maintain.
- The startup team can point to clearer execution, better documentation, stronger collaboration, and practical improvements connected to this role.
How to apply
- Email your resume/CV to Jobs@MetasoftCapital.com and include JID: 4094228 in the email subject line and at the beginning of your message.
- Use a subject line similar to: JID: 4094228 - Data Scientist - Resume/CV.
- Briefly explain why this role, a remote startup environment, and the relevant industry or function fit your experience, interests, and learning goals.
- If useful for the role, include links to your portfolio, GitHub, LinkedIn, website, writing samples, case studies, dashboards, design files, campaigns, or other work examples.