Title
Prompt Engineer
Quick Summary
Candescent AI is hiring a Prompt Engineer to design, test, and productionize prompts and prompt-driven pipelines for our search, chat, and automation features. You will partner with product managers, data scientists, and platform engineers to translate business goals into measurable language model behaviors, with a strong focus on reliability, safety, and cost efficiency. We welcome strong graduates and early-career engineers with demonstrable projects in NLP, retrieval, and evaluation.
Project Category or Industry
Artificial intelligence for SaaS, knowledge management, and enterprise productivity.
Type
Full-time employment.
Experience Level
Entry to mid-level, with mentorship and clear paths for advancement; experienced applicants are also encouraged.
Duration
Permanent role.
Location
Remote-first with optional collaboration hubs in Toronto and Singapore. Maintain at least four hours of overlap with teams operating between UTC-5 and UTC+8.
Salary
USD 85,000–120,000 base depending on location and experience, plus benefits and an annual performance bonus.
Payment Mode
Monthly payroll where supported; compliant contractor arrangements available in select countries.
Hiring Company Name
Candescent AI
Required Skills or Tools
Proficiency in Python and strong written communication; practical understanding of prompt design patterns, retrieval-augmented generation, evaluation metrics, and safety guardrails; familiarity with model APIs and vector databases; ability to ship well-tested services with clear documentation.
Project Description
Candescent AI builds language-centric features that help teams search, summarize, and act on unstructured information. The Prompt Engineer role focuses on turning product requirements into robust prompt strategies and lightweight adapters that deliver predictable outputs across a range of tasks—from structured extraction to conversational flows and workflow automation. The work blends rapid prototyping with disciplined evaluation, red-teaming, and observability so that behaviors remain consistent as data and models evolve.
Core Responsibilities and Expected Deliverables
Design, iterate, and version prompts and templates for classification, summarization, extraction, conversational flows, and tool use.
Build retrieval-augmented pipelines including chunking strategies, embeddings, ranking, and query rewriting.
Create offline and online evaluation harnesses with golden datasets, regression tests, and bias/safety checks.
Ship prompt-driven services as well-tested APIs and workflows; define SLIs/SLOs for quality, latency, and cost.
Monitor behavior in production, analyze failure modes, and deliver continuous improvements backed by metrics.
Produce concise documentation, runbooks, and experiment reports for cross-functional stakeholders.
Required Experience and Preferred Qualifications
Solid Python skills and software engineering fundamentals, including testing and code review.
Experience with at least one modern LLM stack (Hugging Face Transformers, OpenAI/Anthropic/Google APIs, or vLLM) and vector search (FAISS, pgvector, or similar).
Working knowledge of evaluation techniques, prompt libraries, and safety tooling.
Preferred: familiarity with DSPy or Guidance, LangChain or LlamaIndex, prompt-foo or Ragas, and basic statistics for A/B testing.
Evidence of impact through internships, open-source, coursework, or personal projects will be valued.
Tools or Platforms to Be Used
Modeling and experimentation: Python, PyTorch or TensorFlow as needed, Hugging Face, MLflow or Weights & Biases.
Retrieval and storage: Postgres/pgvector or FAISS, Elasticsearch or OpenSearch, object storage for artifacts.
Services and infrastructure: FastAPI, Docker, Kubernetes, GitHub Actions, AWS or GCP, Terraform in collaboration with platform teams.
Observability and safety: Prometheus, Grafana, OpenTelemetry-compatible logging, and policy engines for content filtering.
Language Requirement
Professional English is required. Additional languages are helpful for cross-regional collaboration.
Communication Style
Written-first culture using GitHub issues and pull requests for design and reviews; Slack for daily coordination; Zoom for stand-ups, demos, and incident reviews. Clear documentation is expected for all changes.
Time Commitment or Working Window
Standard 40 hours per week with flexible scheduling. Maintain a predictable daily block that overlaps at least four hours with the core team between 09:00 and 17:00 in your local time.
Payment Terms
Salary paid monthly via payroll. For contractors, invoices are processed on net-30 terms upon acceptance of deliverables and timesheets.
Evaluation Criteria
Portfolio and code samples demonstrating prompt design, evaluation rigor, and thoughtful trade-offs.
Practical exercise focused on building an evaluated prompt pipeline with retrieval and guardrails.
Technical interview on observability, safety, and cost/performance optimization.
Final conversation on product sense, collaboration, and communication.
References may be requested.
Other Requirements
New hires sign a confidentiality agreement and adhere to security and data-handling policies. Light time-tracking may be used for distributed coordination. Occasional shared on-call for language services may be required.
About Candescent AI
Candescent AI is a privately held software company focused on AI-enabled productivity tools for knowledge-heavy teams in finance, healthcare, and professional services. Headquartered in Toronto with a distributed team across North America, Europe, and APAC, we combine rigorous engineering with applied research to deliver reliable language systems. Learn more at https://www.candescent.ai and reach our hiring team at careers@candescent.ai.