Senior Software Engineer - AI Platform

About Ascend Money Ascend Money is a leading fintech company providing innovative payment and financial services across 7 countries in the Southeast Asian Region. Established in 2013, Ascend Money became Thailand’s first fintech unicorn in 2021. Its flagship service TrueMoney today has become the most popular digital financial application that enables ease of payments and convenient financial lifestyle. TrueMoney’s extensive agent network as well as offline and online payment services also enable millions of users across the region to access innovative financial services, leading them to better lives. About the Role: We are building an enterprise AI platform that powers intelligent products across the organization — including conversational AI chatbots, voice AI agents, RAG-based knowledge services, and document intelligence (OCR and image classification for financial documents). As a Senior Software Engineer, you will design and build the core platform services that make it fast and safe for product teams to ship AI-powered features: knowledge base APIs, agent orchestration, model integration layers, and the pipelines that connect them. You will own services end-to-end — from architecture and API design through implementation, testing, deployment, and production operations — and help set the technical direction for how AI capabilities are built and reused across teams. Key Responsibilities: Design, build, and operate backend services for the AI platform, including knowledge base (RAG) services, ingestion pipelines, and vector database integrationsDevelop and maintain LLM-powered applications: conversational chatbots, voice AI agents with tool calling, and document extraction (OCR / image classification) servicesIntegrate and evaluate foundation models and AI platforms (e.g., AWS Bedrock, Google Gemini, agent frameworks such as LangChain/LangGraph), and build abstraction layers so product teams can adopt them easilyDesign clean, versioned APIs for platform capabilities (e.g., knowledge base versioning, document management, release workflows) and support internal teams migrating onto themLead migrations of existing AI services onto the new platform, ensuring reliability and minimal disruptionBuild real-time integrations where needed (e.g., WebSocket-based voice/media streaming for voice AI use cases)Own quality: automated testing, observability (logging, metrics, tracing), performance tuning, and cost optimization for LLM workloadsContribute to technical evaluations and architecture decisions; write clear design docs and documentation (Confluence)Mentor mid-level engineers and review code, raising the engineering bar across the team Qualifications 5+ years of professional software engineering experience, with strong backend skills in Python (FastAPI or similar) and/or Node.js/TypeScriptProven experience designing and operating production microservices and REST APIs on cloud infrastructure (AWS preferred)Hands-on experience building LLM applications: prompt engineering, tool/function calling, RAG pipelines, embeddings, and vector databasesExperience with at least one major AI/model platform: AWS Bedrock, Google Vertex AI / Gemini, OpenAI, or Anthropic APIsSolid engineering fundamentals: system design, data modeling, testing, CI/CD (e.g., Bitbucket Pipelines), containerization (Docker/Kubernetes)Ability to lead a workstream independently — from ambiguous requirements to a shipped, documented, monitored serviceGood written and spoken English for design docs and cross-team communication Preferrable AI Skills: Experience with agent frameworks (LangChain, LangGraph) or building custom agent orchestrationReal-time systems experience: WebSockets, media/audio streaming, telephony or voice AI integrationsDocument AI experience: OCR, structured extraction from financial documents (payment slips, QR codes, bank statements)Experience with MLOps/LLMOps: evaluation harnesses, model versioning, guardrails, cost/latency monitoringFamiliarity with the Thai fintech/payments domain and Thai-language NLP challengesExperience with MLOps/LLMOps: evaluation harnesses, model versioning, guardrails, cost/latency monitoringKnowledge of security and data governance practices for AI systems (PII handling, compliance)