Ascensus is the leading independent technology and service platform powering savings plans across America, providing products and expertise that help nearly 16 million people save for a better today and tomorrow.
Ascensus does not provide visa sponsorship now or in the future and does not hire candidates working under OPT.
Section 1: Position Summary
- Associate Level (0-2 years experience /New Grads)
- Mid Level (3-5 years experience)
- Senior Level (5-15 years experience)
- SDET (3+ years experience)
Work at the cutting edge of enterprise AI, solving complex challenges and improving how intelligent systems perform at scale. In this role, you'll troubleshoot AI behavior, partner with other engineers to fix root causes, and continuously enhance platform performance across LLMs, agent workflows, and backend systems. This isn't traditional support-you'll play a hands‑on role in shaping how AI is built, used, and experienced across the organization.
Section 2: Job Functions, Essential Duties and Responsibilities
- Technical Support & Triage
- Investigate and resolve issues reported by users of Agent Assist
- Reproduce issues and trace failures across:
- LLM responses and prompts
- Agent workflows and orchestration
- Tool/MCP calls and backend integrations (e.g., Salesforce and internal enterprise applications)
- Analyze logs, traces, and telemetry (e.g., Langfuse, NewRelic logs)
- Clearly document findings and partner with engineering on resolution leveraging Azure DevOps
- AI System Debugging & Improvement
- Identify whether issues stem from prompting, retrieval (RAG), tool selection, or backend systems
- Propose improvements to agents, skills, prompts, and workflows
- Use AI-powered tools (e.g., Cursor) to accelerate debugging, analysis, and solutioning
- Contribute to continuous improvement of system reliability and response quality
- User Enablement & Evangelism
- Train internal users on how to effectively use Agent Assist
- Run targeted demos, onboarding sessions, and office hours
- Act as a trusted advisor to client‑facing associates
- Drive adoption by helping users understand both capabilities and limitations
- Cross‑Functional Collaboration
- Partner closely with engineering, product, and AI teams
- Translate user issues into clear, actionable tickets and improvements
- Identify recurring patterns and elevate systemic issues
- General
- Responsible for protecting, securing, and proper handling of all confidential data held by Ascensus to ensure against unauthorized access, improper transmission, and/or unapproved disclosure of information that could result in harm to Ascensus or our clients.
- At Ascensus we are guided by our Core Values of People Matter, Quality First and Integrity Always®. They inspire us every day to prioritize an environment of respect for those we serve and one another and should be visible in your actions on a day-to-day.
Section 3: Experience, Skills, Knowledge Requirements
- Up to 3 years' experience in a Software Engineering, Software Support, QA, SDET, or engineering-adjacent role (including internships, co‑ops, or academic projects)
- Bachelor's degree in Computer Science, Computer Information Systems, Business Information Systems, or a related technical field
- Demonstrated ability to troubleshoot and debug technical issues, focusing on root cause analysis
- Foundational experience or exposure to:
- API testing (Postman or similar)
- Reading logs and debugging application behavior
- Basic SQL or data inspection
- Exposure to or strong interest in AI/LLM-based systems (prompting, chatbots, RAG, or agents) through coursework, projects, or hands‑on experimentation
- Experience using or willingness to learn AI-powered development tools (e.g., Cursor, Claude Code)
- Strong communication skills with the ability to translate technical issues into clear explanations for non‑technical users
- Demonstrated curiosity and ability to learn quickly in a fast‑evolving technical environment
- Experience supporting or building AI/agent-based applications is preferred
- Familiarity with the following is preferred:
- Retrieval‑Augmented Generation (RAG)
- Tool/agent orchestration
- Prompt engineering
The national av