AI Engineer- SaaS Platform

Accountabilities:

• • Maintain and optimize LLM- and VLM-powered services supporting content generation, compliance scoring, campaign testing, and other AI-driven workflows.
  • Develop and maintain production-grade Python services using Flask and FastAPI, ensuring high availability, reliability, and low latency.
  • Manage asynchronous AI workflows and pipeline orchestration using Dramatiq or comparable background job technologies.
  • Deploy, monitor, troubleshoot, and optimize Uvicorn/Gunicorn-based services in production environments.
  • Integrate LLM/VLM routing and fallback solutions such as OpenRouter to balance model quality, latency, reliability, and infrastructure costs.
  • Design and refine prompt engineering strategies that improve reliability, contextual relevance, consistency, and compliance of AI outputs.
  • Build and maintain structured AI feedback and evaluation pipelines incorporating human-in-the-loop scoring, automated quality checks, and continuous improvement mechanisms.
  • Design, secure, version, document, and maintain REST APIs supporting AI services and integrations.
  • Collaborate with backend and frontend engineering teams to ensure microservices remain scalable, maintainable, and aligned with broader platform architecture.
  • Monitor production metrics including token consumption, latency, error rates, uptime, and service health, proactively identifying and resolving performance issues.
  • Support Dockerized deployments, CI/CD workflows, logging, monitoring, error handling, and production reliability practices across AI services.

  Requirements:

  • 3–5 years of professional experience as an AI Engineer, Python Backend Engineer, or in a closely related role working with production systems.
  • Strong Python programming skills and experience contributing to production-grade codebases.
  • Hands-on experience with Flask and familiarity with FastAPI is highly valuable.
  • Experience deploying and operating applications with Uvicorn and/or Gunicorn.
  • Experience with asynchronous job queues such as Dramatiq, Celery, RQ, or equivalent technologies.
  • Practical experience working with LLMs and VLMs, including prompt engineering, model evaluation, and ideally fine-tuning.
  • Familiarity with LLM/VLM routing, fallback, and orchestration platforms such as OpenRouter or equivalent solutions.
  • Strong understanding of microservice architecture and experience designing, maintaining, or scaling distributed services.
  • Solid REST API development skills, including authentication, rate limiting, versioning, documentation, and secure endpoint design.
  • Experience with Dockerized deployments, CI/CD pipelines, application logging, monitoring, debugging, and production error handling.
  • Demonstrated ability to build structured AI evaluation and feedback systems that continuously measure and improve model performance.
  • Experience with AWS or GCP for deploying, monitoring, and scaling production workloads is preferred.
  • Experience working with SaaS platforms, AI-first products, or production LLM/VLM integrations is highly valued.
  • Demonstrated ability to maintain and improve AI pipelines in production rather than working primarily with prototypes or experimental systems.
  • Strong problem-solving, communication, collaboration, and ownership skills, with the ability to work effectively in a fast-moving technical environment.

  Benefits:

  • Fully remote position based in India.
  • Full-time employment within a SaaS and AI-focused product environment.
  • Opportunity to work hands-on with production LLM and VLM systems.
  • Exposure to modern AI infrastructure, model routing, microservices, APIs, asynchronous processing, and cloud technologies.
  • Opportunity to contribute directly to AI reliability, performance, evaluation, and continuous improvement.
  • Collaborative work with backend and frontend engineering teams on scalable platform architecture.
  • Opportunity to develop deeper expertise in production AI engineering and SaaS platform development.