Data Scientist Spring Intern— Agentic AI at SS&C… · Waltham

Summary: SS&C Technologies is a leading financial services and healthcare technology company headquartered in Windsor, Connecticut. They are seeking a Data Scientist Spring Intern to build and iterate AI agents, develop AI-powered coding terminals, and experiment with open-source LLMs to enhance agent capabilities and performance. Responsibilities: - Build and iterate AI agents end-to-end — from defining goals, personas, and constraints to wiring LLM reasoning with tool execution - Implement planning strategies including chain-of-thought, ReAct loops, and hierarchical task decomposition to enable agents to solve multi-step problems autonomously - Design and refine system prompts, few-shot examples, and guardrails that shape agent behavior, tone, and decision-making boundaries - Work extensively inside AI-powered coding terminals (OpenCode, DeepAgent) as both a user and a builder — understanding how these tools orchestrate LLM calls, file edits, and shell commands - Contribute to the development of custom coding agent workflows that automate code generation, review, refactoring, and testing tasks - Evaluate and benchmark terminal agent behaviors: accuracy of code edits, hallucination rates, context utilization, and multi-file reasoning - Build and extend MCP servers (using FastMCP and similar frameworks) that expose internal tools, databases, and APIs as structured capabilities for agents - Design tool schemas, descriptions, and invocation patterns that LLMs can reliably discover and call - Integrate agents with external services — REST APIs, vector stores, graph databases, and internal SDKs — through well-defined MCP interfaces - Experiment with a range of open-source LLMs (Qwen, DeepSeek, finetune domain-specific models) to evaluate reasoning quality, latency, cost, and tool-use reliability - Explore inference optimizations such as speculative decoding, constraint decoding, structured outputs, and router-mode orchestration - Build and run evaluation pipelines to measure retrieval accuracy, tool-selection precision, hallucination rate, and end-to-end task completion - Integrate agents with a broad ecosystem of external systems: vector stores (PgVector, Milvus), relational and graph databases, REST APIs, and internal microservices, all managed through secure, least-privilege access patterns - Design and test memory architectures — short-term (conversation context), long-term (vector-stored interaction history), and episodic (task-specific recall) — to improve agent continuity and personalization Required Qualifications: - Build and iterate AI agents end-to-end — from defining goals, personas, and constraints to wiring LLM reasoning with tool execution - Implement planning strategies including chain-of-thought, ReAct loops, and hierarchical task decomposition to enable agents to solve multi-step problems autonomously - Design and refine system prompts, few-shot examples, and guardrails that shape agent behavior, tone, and decision-making boundaries - Work extensively inside AI-powered coding terminals (OpenCode, DeepAgent) as both a user and a builder - Contribute to the development of custom coding agent workflows that automate code generation, review, refactoring, and testing tasks - Evaluate and benchmark terminal agent behaviors: accuracy of code edits, hallucination rates, context utilization, and multi-file reasoning - Build and extend MCP servers (using FastMCP and similar frameworks) that expose internal tools, databases, and APIs as structured capabilities for agents - Design tool schemas, descriptions, and invocation patterns that LLMs can reliably discover and call - Integrate agents with external services — REST APIs, vector stores, graph databases, and internal SDKs — through well-defined MCP interfaces - Experiment with a range of open-source LLMs (Qwen, DeepSeek, finetune domain-specific models) to evaluate reasoning quality, latency, cost, and tool-use reliability - Explore inference optimizations such as speculative decoding, constraint decoding, structured outputs, and router-mode orchestration - Build and run evaluation pipelines to measure retrieval accuracy, tool-selection precision, hallucination rate, and end-to-end task completion - Integrate agents with a broad ecosystem of external systems: vector stores (PgVector, Milvus), relational and graph databases, REST APIs, and internal microservices - Design and test memory architectures — short-term (conversation context), long-term (vector-stored interaction history), and episodic (task-specific recall) — to improve agent continuity and personalization Required Skills: AI agent design, Large Language Models (LLMs), Chain-of-thought planning, ReAct loops, Hierarchical task decomposition, Prompt engineering, AI coding terminals, Code generation, Code review, Code refactoring, Code testing, Model Context Protocol (MCP), FastMCP framework, REST APIs, Vector stores, Graph databases, SDK integration, Open-source LLM experimentation, Inference optimization, Evaluation pipelines, Memory architectures, Relational databases, Microservices Benefits: Interns are eligible to participate in SS&C's medical plan, Health Savings Account (HSA)