Artificial Intelligence Engineer

Total CollectR (product of Total AI Systems Inc.) is on a mission to revolutionize debt collection with AI.

We are looking for an Applied AI Engineer to design, build, and ship AI-powered features using modern large-language-model (LLM) technologies. You will work at the intersection of software engineering, machine learning, and product development, turning real business problems into reliable, measurable AI solutions.

This role is hands-on: youll prototype quickly, build production-grade systems, evaluate model quality, and collaborate closely with product, data, design, and domain experts.

What Youll Do

AI Feature Development

• Build end-to-end LLM-powered applications (retrieval-augmented generation, agents, classifiers, summarizers, etc.).
• Design and optimize prompts, system instructions, and model interaction flows.
• Implement retrieval pipelines (vector search, embeddings, chunking, metadata filtering).
• Integrate AI components into backend services, APIs, and product features.

Model & Data Integration

• Select, tune, and integrate appropriate models (open-source, hosted APIs, fine-tuned models).
• Manage embeddings, vector stores, and data pipelines required for RAG and model inputs.
• Ensure data is handled securely (PII redaction, access controls, auditability).

Evaluation & Quality

• Build evaluation suites for accuracy, hallucination risk, safety, and reliability.
• Run A/B tests, benchmark models, and tune performance/cost tradeoffs.
• Instrument applications with robust telemetry (latency, tokens, error modes, user feedback loops).

Production Engineering

• Develop scalable, resilient inference services with caching, batching, streaming, and model fallbacks.
• Work with MLOps/Platform teams to deploy and monitor AI services in production.
• Create guardrails, validation layers, and safety protections around model output.
• Write clean, testable code and documentation for maintainable AI systems.

Collaboration

• Work with PMs to understand user needs and define success metrics.
• Partner with designers to craft intuitive, safe AI user experiences.
• Collaborate with domain experts to build ground-truth datasets and eval criteria.
• Participate in design reviews, architecture discussions, and roadmap planning.

What Youll Bring

Required

• Strong software engineering background (Python, TypeScript/Node, or similar).
• Experience building production-grade systems or backend services.
• Hands-on experience with modern LLMs (OpenAI, Anthropic, open-source models, etc.).
• Familiarity with embeddings, vector databases (FAISS, Pinecone, Weaviate, etc.), and RAG patterns.
• Ability to design prompts, evaluate outputs, and iterate systematically.
• Understanding of API design, cloud infrastructure, and containerization (AWS/GCP/Azure, Docker, etc.).
• Strong problem-solving skills and a product-focused mindset.

Nice to Have

• Experience with agentic systems, tool-calling, or multi-step pipelines.
• Experience with model fine-tuning, LoRA, or supervised instruction tuning.
• Background in data engineering or MLOps (Airflow, Dagster, CI/CD for ML, feature stores).
• Experience with observability (OpenTelemetry, logging/tracing pipelines).
• Knowledge of safety/guardrails frameworks (content filtering, adversarial testing, RAI principles).
• Experience building eval suites or working with labeled datasets.

Who You Are

• You are curious, scrappy, and comfortable working on ambiguous problems.
• You move fast but care deeply about reliability and quality.
• You think in terms of user value, not just model APIs.
• You enjoy learning new techniques and working across disciplines.
• You want to shape how AI is applied in real-world products.

Success in This Role Looks Like

• Shipping AI features that create measurable improvements for users or operations.
• Reducing hallucinations and improving reliability through structured evals and guardrails.
• Designing systems that scale with usage while controlling cost and latency.
• Partnering effectively across product, engineering, and data teams.
• Establishing reusable best practices, prompts, pipelines, and tooling.