Frost & Sullivan este o companie globală de consultanță și cercetare de piață care deservește organizații din domeniul sănătății, marketingului și din alte industrii. Echipa sa combină cercetarea sectorială, informațiile despre piață, analiza datelor — inclusiv motorul FROST AI — și serviciile de consultanță strategică pentru a ajuta companiile, investitorii și organizațiile din sectorul public să evalueze oportunități, să definească strategii de inovare, să planifice lansări de produse, să evalueze decizii privind fuziuni și achiziții și să își consolideze planurile de intrare pe piață. Compania reunește, de asemenea, lideri de afaceri prin evenimente și consilii de conducere, conectând cercetarea cu planificarea practică a creșterii în diverse sectoare și regiuni geografice.
AI Model Trainer, Business Technology – Singapore Hybrid
Train and evaluate Frost & Sullivan’s MetaBrain by converting research and business expertise into structured evaluation assets, rules, and tested workflow improvements. The role combines expert review, quality measurement, controlled iteration, and— for the technology track—Python-based evaluation and dataset management.
Descriere
- Teach and assess MetaBrain’s use of approved research methods and business logic
- Translate subject-matter expertise into structured examples, rules, and review criteria
- Convert advisor feedback into controlled, testable model improvements
- Organize authorized research, taxonomies, decision rules, examples, and counterexamples
- Develop sourced reference answers and document scope, dates, limitations, and permitted use
- Create rubrics covering factual support, numerical consistency, relevance, completeness, and uncertainty
- Conduct blind comparisons and record errors, reviewer disagreements, and adjudication outcomes
- Maintain separation between development material and held-out evaluation cases
- Configure prompts, retrieval settings, and business rules in approved design and tuning environments
- Convert advisor corrections into version-controlled change requests
- Retest each update and maintain a documented rollback record
- Collaborate with technical trainers on experiment design and failure analysis
- For the technology track, develop Python evaluation scripts, manage dataset versions, and support justified fine-tuning experiments
Cerințe
- At least three years of substantial research, consulting, or knowledge-quality experience, with equivalent accomplishments considered
- Demonstrated ability in evidence assessment, business writing, numerical validation, and methodological precision
- Foundational knowledge of AI or large language models through formal study or certification, plus hands-on experience testing or improving AI outputs
- Experience with peer review, analyst coaching, survey coding, taxonomy development, quality assurance, low-code automation, or structured knowledge-base maintenance is preferred
- SQL or Python experience is useful for business-track candidates
- Python evaluation scripting is required for the technology track