Interested in this AI Software Engineer role at The Hartford?
Apply Now →Skills & Technologies
About This Role
Job Details
---------------
Location:
Columbus, OH
Category:
Information Technology
Employment Type:
Full time
Job Ref:
R2626287\-174
Senior Staff Software Engineer \- IE07HE
We’re determined to make a difference and are proud to be an insurance company that goes well beyond coverages and policies. Working here means having every opportunity to achieve your goals – and to help others accomplish theirs, too. Join our team as we help shape the future.
This requisition hires Senior AI Engineers who will:
Design and deliver production‑grade Agentic AI systems using Google ADK, Anthropic MCP, LangGraph/LangChain, and modern Agentic protocols. Build secure, scalable AI platform capabilities with strong engineering fundamentals in Python/Typescript, Terraform, and GCP. Enable enterprise adoption of AI by creating reusable frameworks, APIs, and platform capabilities aligned with engineering standards, compliance needs, and modern cloud patterns.
Overview
The Senior AI Engineer will architect, build, and operationalize advanced AI and multi\-agent solutions leveraging RAG, GraphRAG, Agentic AI frameworks, and enterprise‑grade cloud engineering.
A key requirement is robust, practical experience implementing MCP and ADK Agentic Protocols, with a solid understanding of:
- Agent memory
- Session and context lifecycle management
- Tooling interfaces
- Secure capability boundaries
- Permissions and role enforcement
Additionally, candidates must have hands\-on experience with AlloyDB’s AI/Agentic capabilities—including vector indexing, embedding support, and tight integration with Vertex AI—as well as strong fundamentals in PostgreSQL / Postgres RDS for building retrieval systems, agent memory stores, and structured context\-management layers.
The engineer must demonstrate strong foundational engineering skills in Python or Typescript, IaC (Terraform), DevOps pipelines, and secure distributed system design using GCP services such as Vertex AI, Cloud Run, Cloud Storage, and AlloyDB.
The role additionally requires deep, hands\-on experience building and extending agent harnesses—the runtime scaffolding that orchestrates the agent execution loop, tool invocation, dynamic context\-window assembly, sub\-agent delegation, and guardrail and permission enforcement—together with production expertise in LangChain and LangGraph.
Fluency in spec\-driven, agentic development frameworks such as GitHub Spec\-Kit, OpenSpec, and BMAD\-METHOD, used to translate intent into executable specifications and orchestrate AI\-assisted delivery at enterprise scale.
Responsibilities
AI/Agentic System Architecture \& Development
- Design and implement Agentic AI solutions using Google ADK, LangGraph, LangChain, and Agent Engine.
- Build and extend agent harnesses, implementing the agent execution loop, tool\-call orchestration, dynamic prompt and context assembly, sub\-agent delegation, streaming, token\-budget management, and hook and guardrail enforcement.
- Engineer advanced LangChain and LangGraph orchestration, including LCEL chains, stateful graphs, checkpointing, human\-in\-the\-loop workflows, memory, retrievers, callbacks, and LangSmith tracing and evaluation.
- Build advanced RAG and GraphRAG pipelines, vector retrieval systems, and knowledge‑graph–augmented reasoning.
Implement MCP\-compliant agents with capability registration, secure tool invocation, memory storage, and session state management.
- Apply deep knowledge of Agentic Protocol design (ADK \& MCP), such as:
+ Agent memory and conversation state
+ Tool authorization
+ Multi‑step workflows and orchestration
+ Session boundary and identity controls
- Leverage AlloyDB and PostgreSQL/RDS for:
+ Vector storage and hybrid search
+ Agent memory persistence, session management, and state recovery
+ Structured prompt scaffolding and fact retrieval
+ ACID‑compliant transactional reasoning layers
- Develop scalable AI microservices using Python/Typescript, Cloud Run, Vertex AI, and event\-driven components.
- Optimize model inference, retrieval latency, and overall system performance.
Spec\-Driven \& Agentic Development
- Drive spec\-driven development (SDD) using frameworks such as GitHub Spec\-Kit, OpenSpec, and BMAD\-METHOD, translating product intent into executable specifications, plans, and agent\-ready task breakdowns.
- Establish specification\-first review gates and living change proposals that align human engineers and AI agents before implementation begins.
Security, Governance \& Session Management
- Implement enterprise\-grade security for agents including:
+ OAuth and SSO flows
+ IAM roles, service accounts, least‑privilege design
+ Secure MCP tool access, command permissioning, and input validation
- Architect safe session‑based AI interactions with proper expiration, auditing, and context isolation.
- Ensure compliance with enterprise governance, Responsible AI requirements, and platform guardrails.
Platform Engineering, IaC \& DevOps
- Use Terraform to build GCP infrastructure for AI workloads, vector stores, knowledge graphs, and orchestration services.
- Build CI/CD pipelines for model deployments and agent lifecycle automation.
- Implement observability, monitoring, and logging for AI service health.
Innovation \& Collaboration
- Evaluate emerging tools and frameworks—including Claude Code, GitHub Copilot, AWS Kiro, GitHub Spec\-Kit, OpenSpec, and BMAD\-METHOD—and integrate them into engineering workflows.
- Partner with architects, data engineers, and platform teams to implement cross‑domain AI capabilities.
- Document architecture patterns, reusable code modules, and standards for MCP/Agentic development.
Qualifications
Experience
- 6–8 years in software engineering, including 2\+ years in GenAI, multi\-agent, or LLM systems.
- Proven delivery of at least one production‑grade AI or Agentic system, preferably involving RAG or GraphRAG.
Technical Expertise
Core Engineering
- Strong engineering fundamentals in Python and/or Typescript.
Agentic AI \& Protocols
- Deep, practical experience with:
+ MCP (Model Context Protocol) — tools, capabilities, memory, session orchestration, security
+ Google ADK Agentic Protocols — agents, workflows, context management
+ LangChain \& LangGraph — LCEL chains, agents, tools, memory, retrievers, stateful graph orchestration, checkpointing, human\-in\-the\-loop control, and LangSmith tracing and evaluation
+ Agent harness engineering — agent execution loops, tool\-call orchestration, context and prompt assembly, sub\-agent delegation, streaming, token\-budget management, and hook and guardrail enforcement
Spec\-Driven \& Agentic Development Frameworks
- Hands\-on experience with spec\-driven development (SDD) workflows and tooling, including GitHub Spec\-Kit (specify, plan, tasks, implement), OpenSpec (change proposals and living specifications), and BMAD\-METHOD (agentic planning with specialized agent roles)
- Proven ability to decompose product intent into executable specifications, structured plans, and agent\-ready task breakdowns that align human and AI contributors before code is written
- Familiarity with greenfield and brownfield delivery driven by multi\-agent planning, context engineering, and specification\-first review gates
Databases \& Agent Memory Stores
- Hands‑on experience with AlloyDB, including:
+ Vector indexing / pgvector
+ AI inference acceleration and Vertex AI integration
+ Building agent memory and retrieval layers
+ Transactional context management for Agentic systems
- Strong PostgreSQL/Postgres RDS fundamentals, including:
+ Schema design for knowledge retrieval
+ Query optimization
+ Hybrid search patterns
+ Durable storage for AI session and memory state
Cloud \& Platform Skills
- Experience with:
+ Vertex AI (Model Garden, Embeddings, Vector Search, Generative AI APIs)
+ GCP Cloud Run, AlloyDB, Cloud Storage, Secret Manager
+ Terraform / IaC
+ CI/CD automation, containerization, environment provisioning
+ OAuth, SSO, IAM roles/policies, service account management
Additional
- Experience with AI coding tools (Claude Code, GitHub Copilot, AWS Kiro).
- Strong understanding of LLM safety, governance, context window management, and prompt engineering.
Preferred Certifications
- GCP Professional Cloud Architect
- GCP Professional Machine Learning Engineer
Education
- Bachelor’s or Master’s in Computer Science, Engineering, or related field.
This role will have a Hybrid work schedule, with the expectation of working in an office (Columbus, OH, Chicago, IL, Hartford, CT or Charlotte, NC) 3 days a week (Tuesday through Thursday). Candidates must be authorized to work in the US without company sponsorship. The company will not support the STEM OPT I\-983 Training Plan endorsement for this position.
Compensation
The listed annualized base pay range is primarily based on analysis of similar positions in the external market. Actual base pay could vary and may be above or below the listed range based on factors including but not limited to performance, proficiency and demonstration of competencies required for the role. The base pay is just one component of The Hartford’s total compensation package for employees. Other rewards may include short\-term or annual bonuses, long\-term incentives, and on\-the\-spot recognition. The annualized base pay range for this role is:
$127,600 \- $191,400
Equal Opportunity Employer/Sex/Race/Color/Veterans/Disability/Sexual Orientation/Gender Identity or Expression/Religion/Age
Salary Context
This $127K-$191K range is below the median for AI Software Engineer roles in our dataset (median: $185K across 231 roles with salary data).
Role Details
About This Role
AI Software Engineers build the applications and systems that AI models run inside. They own the API layers, data pipelines, frontend integrations, and infrastructure that turn a model into a product users interact with. Every AI company needs engineers who can build the software around the AI.
The challenge is building reliable systems around inherently unreliable components. Models are probabilistic. They'll give different answers to the same question. They hallucinate. They're slow. They're expensive. Your job is to build an application layer that handles all of this gracefully while delivering a product that users trust and enjoy.
Across the 4,317 AI roles we're tracking, AI Software Engineer positions make up 7% of the market. At The Hartford, this role fits into their broader AI and engineering organization.
AI Software Engineer roles are among the most numerous in the AI job market. Every company deploying AI needs software engineers who understand AI integration patterns. The demand is broad, spanning startups to enterprises, across every industry adopting AI capabilities.
What the Work Looks Like
A typical week includes: building API endpoints that serve model inference with caching and fallback logic, designing the data pipeline that feeds context to a RAG system, implementing streaming responses in the frontend, debugging a race condition in the async inference pipeline, and optimizing database queries for the vector search layer. It's full-stack engineering with AI at the center.
AI Software Engineer roles are among the most numerous in the AI job market. Every company deploying AI needs software engineers who understand AI integration patterns. The demand is broad, spanning startups to enterprises, across every industry adopting AI capabilities.
Skills Required
Full-stack engineering skills with AI integration experience. Python and TypeScript are the most common requirements. You'll need to understand API design, database architecture, and how to build reliable systems around probabilistic outputs. Experience with streaming, async processing, and caching patterns is increasingly important as real-time AI applications proliferate.
Knowledge of vector databases, embedding APIs, and LLM integration patterns (function calling, structured outputs, retry logic) differentiates AI software engineers from general software engineers. Understanding cost optimization (caching strategies, model routing, batched inference) is valuable since inference costs can dominate application economics.
Strong postings describe the product you'll be building, the AI integration patterns you'll work with, and the scale requirements. Look for companies that have existing AI features and need engineers to improve and expand them, not companies that are 'planning to add AI' someday.
Compensation Benchmarks
AI Software Engineer roles pay a median of $218,500 based on 729 positions with disclosed compensation. Senior-level AI roles across all categories have a median of $227,400. This role's midpoint ($159K) sits 27% below the category median. Disclosed range: $127K to $191K.
Across all AI roles, the market median is $215,000. Top-quartile compensation starts at $266,300. The 90th percentile reaches $320,790. For comparison, the highest-paying categories include AI Safety ($287,500) and Research Engineer ($272,100). By seniority level: Entry: $110,000; Mid: $194,400; Senior: $227,400; Director: $274,554; VP: $241,000.
The Hartford AI Hiring
The Hartford has 4 open AI roles right now. They're hiring across AI Software Engineer, AI/ML Engineer. Positions span Columbus, OH, US, Chicago, IL, US. Compensation range: $175K - $273K.
Location Context
Across all AI roles, 15% (635 positions) offer remote work, while 3,657 require on-site attendance. Top AI hiring metros: New York (1,650 roles, $220,000 median); San Francisco (1,335 roles, $265,000 median); Los Angeles (708 roles, $214,112 median).
Career Path
Common paths into AI Software Engineer roles include Software Engineer, Full-Stack Developer, Backend Engineer.
From here, career progression typically leads toward Staff Engineer, AI Architect, Engineering Manager.
If you're a software engineer, you're already 80% there. Learn the AI integration patterns: RAG, streaming inference, function calling, structured outputs. Build a project that demonstrates you can wrap an AI model in a production-quality application with proper error handling, caching, and user experience. That's the portfolio piece that gets you hired.
What to Expect in Interviews
Technical screens look like standard software engineering interviews with an AI twist. Expect system design questions about building reliable applications around probabilistic models: handling streaming responses, implementing retry logic for API failures, and designing caching strategies for LLM outputs. Coding rounds test standard algorithms plus practical integration patterns like async processing and rate limiting.
When evaluating opportunities: Strong postings describe the product you'll be building, the AI integration patterns you'll work with, and the scale requirements. Look for companies that have existing AI features and need engineers to improve and expand them, not companies that are 'planning to add AI' someday.
AI Hiring Overview
The AI job market has 4,317 open positions tracked in our dataset. By seniority: 138 entry-level, 2,071 mid-level, 1,655 senior, and 453 leadership roles (Director, VP, C-Level). Remote roles make up 15% of the market (635 positions). The remaining 3,657 roles require on-site or hybrid attendance.
The market median for AI roles is $215,000. Top-quartile compensation starts at $266,300. The 90th percentile reaches $320,790. Highest-paying categories: AI Safety ($287,500 median, 34 roles); Research Engineer ($272,100 median, 227 roles); AI Engineering Manager ($244,000 median, 23 roles).
AI Software Engineer roles are among the most numerous in the AI job market. Every company deploying AI needs software engineers who understand AI integration patterns. The demand is broad, spanning startups to enterprises, across every industry adopting AI capabilities.
The AI Job Market Today
The AI job market spans 4,317 open positions across 15 role categories. The largest categories by volume: AI/ML Engineer (3,004), Data Scientist (345), AI Software Engineer (309). These three account for the majority of open positions, though smaller categories often have higher per-role compensation because of specialized skill requirements.
The seniority mix tells a story about where AI teams are in their maturity. Entry-level roles (138) are outnumbered by mid-level (2,071) and senior (1,655) positions, reflecting that most companies are past the 'build a team from scratch' phase and need experienced engineers who can ship production systems. Leadership roles (Director, VP, C-Level) total 453 positions, representing the bottleneck between technical execution and organizational strategy.
Remote work availability sits at 15% of all AI roles (635 positions), with 3,657 requiring on-site or hybrid attendance. The remote share has stabilized after the post-pandemic correction. Senior and specialized roles (Research Scientist, ML Architect) are more likely to be remote-eligible than entry-level positions, partly because experienced hires have more negotiating power and partly because these roles require less hands-on mentorship.
AI compensation is structured in clear tiers. The market median sits at $215,000. Top-quartile roles start at $266,300, and the 90th percentile reaches $320,790. These figures include base salary with disclosed compensation. Total compensation (including equity, bonuses, and sign-on) runs 20-40% higher at companies that offer those components.
Category matters for compensation. AI Safety roles lead at $287,500 median, while Prompt Engineer roles sit at $145,000. The spread between highest and lowest-paying categories reflects the premium on specialized technical skills versus broader analytical roles.
The most in-demand skills across all AI postings: Python (2,249 postings), Aws (1,224 postings), Azure (938 postings), Rag (915 postings), Gcp (660 postings), Pytorch (640 postings), Prompt Engineering (624 postings), Kubernetes (559 postings). Python dominates, appearing in the vast majority of role descriptions regardless of category. Cloud platform experience (AWS, GCP, Azure) is the second most common requirement. The newer entrants to the top skills list (RAG, vector databases, LLM APIs) reflect the shift from traditional ML toward generative AI applications.
Frequently Asked Questions
Get Weekly AI Career Intelligence
Salary data, skills demand, and market signals from 16,000+ AI job postings. Every Monday.