Newforma provides industry-leading Information Management and collaboration software for the Architecture, Engineering, Construction, and Owner/Operator (AECO) industry. With over 500,000 users in more than 1,500 firms worldwide, our mission is to empower AECO teams by delivering technology solutions that drive better project outcomes throughout the construction project lifecycle. As a portfolio company of Ethos Capital, we're committed to innovation and excellence in connecting project data, processes, and people.
We're looking for a Staff Platform Developer to join our Platform Engineering team and help build the foundation that our AI-powered future runs on. This is a senior, hands-on role that spans three connected areas: the data platform that feeds our AI and analytics, the harness engineering that makes our agentic framework reliable in production, and our agentic software factory, where AI agents help us design, build, test, and ship software.
You'll be a technical anchor for the team. You'll write a lot of code, but you'll also set patterns, raise the bar on engineering quality, and mentor developers as agentic practices spread across the organization. You'll work closely with our Lead Architect, the Platform Engineering team, and the product teams building on top of the platform, all while we complete our transition from Azure to AWS with the support of our strategic AWS partnership.
Newforma manages billions of emails, documents, and project artifacts for construction teams around the world. We've already shipped AI features like Smart Email Filing, and we're building an extensible agentic framework that orchestrates many specialized agents to search, answer questions, and take action for our users. At the same time, we're changing how we build software, moving from AI-assisted coding toward an agentic software factory where product intent goes in and tested pull requests come out.
All of this depends on a strong platform underneath. You'll help us solve three problems at once:
- Data: getting the right data, in the right shape, to the right agents and dashboards, in near real time
- Harness engineering: giving agents the tools, context, guardrails, evaluations, and observability they need to behave reliably at scale
- The agentic factory: building the pipelines, quality gates, and developer tooling that let agents contribute production-grade code safely
Responsibilities For This Role Include
Data Platform & Engineering
- Contribute to the design and build-out of our medallion architecture (bronze, silver, gold layers) on AWS, supporting both AI initiatives and analytics
- Build and operate event-driven data flows (Kafka/MSK, EventBridge, Kinesis, SQS/SNS) that keep search indexes, vector stores, and analytics pipelines in sync with source systems
- Define repeatable patterns for onboarding new datasets into the platform, including change data capture, snapshots, and deltas
- Build indexing and embedding pipelines that feed our vector databases and semantic search capabilities
- Ensure data is governed, secure, and tenant-aware, working with our DevSecOps practice on access controls and compliance
- Partner with our data architect and data engineers on data modeling, quality checks, and lineage
Harness Engineering for Our Agentic Framework
- Build and evolve the agent harness: the runtime, tool interfaces, context management, memory, and orchestration that our agents run on
- Design how agents access Newforma data and capabilities through well-defined tools and APIs, including Model Context Protocol (MCP) servers
- Implement guardrails, permission boundaries, and human-in-the-loop patterns so agents act safely on behalf of users
- Build evaluation frameworks that measure agent quality, catch regressions, and let teams ship agent changes with confidence
- Establish observability for agentic systems: tracing, cost and token tracking, latency, and failure analysis
- Support multi-model strategies (AWS Bedrock, Anthropic, OpenAI) with routing, fallbacks, and cost optimization
- Make it easy for product teams to build new agents on the framework through reusable components, templates, and clear documentation
Agentic Software Factory
- Help build our agentic software factory, the pipeline that takes product intent through specification, agent implementation, automated quality gates, and human review to a merged pull request
- Design and implement the harness that coding agents operate within: repository context, specifications, test scaffolding, and sandboxed execution environments
- Build automated quality gates (tests, static analysis, security scanning, graders) that make AI-generated code trustworthy
- Measure and improve factory throughput, quality, and cost, and identify where human review adds the most value
- Integrate the factory with our CI/CD pipelines, Pulumi-managed infrastructure, and developer workflows
- Help teams adopt agentic coding practices and share what works across the engineering organization
Technical Leadership
- Act as a senior technical voice on the Platform Engineering team, setting patterns and standards others build on
- Lead technical design for cross-cutting platform initiatives and write clear design documents
- Mentor developers through code reviews, pairing, and knowledge sharing
- Collaborate with the Lead Architect on platform direction and architecture decisions
- Contribute to our Azure-to-AWS transition, ensuring new platform capabilities are built AWS-native
- Occasionally support production systems, including participation in an on-call rotation
Required Qualifications For This Role Include
- 8+ years of software development experience, with at least 3 years working on platforms, infrastructure, or data systems used by other engineering teams
- Strong programming skills in Python and at least one of TypeScript/Node.js or C#/.NET
- Hands-on experience building production systems on AWS (Lambda, ECS/EKS, S3, DynamoDB, RDS, Step Functions, EventBridge)
- Experience designing and operating event-driven architectures and streaming data pipelines (Kafka/MSK, Kinesis, or similar)
- Practical data engineering experience: data lakes or lakehouses, ETL/ELT, data modeling, and ideally medallion-style layered architectures
- Hands-on experience building applications with LLMs and agentic frameworks such as LangGraph, LangChain, the Claude Agent SDK, or similar
- Understanding of what makes agents reliable in production: tool design, context management, evaluations, guardrails, and observability
- Daily, serious use of AI coding agents (Claude Code, Cursor, Copilot, or similar) and strong opinions on how to make them productive and safe
- Experience with infrastructure as code and CI/CD pipelines
- Proven ability to lead technical designs across teams and influence without direct authority
- Excellent communication skills and a habit of documenting decisions clearly
- Ability to travel up to 10% for company and team events
Nice To Have Qualifications For This Role Include
- Experience building or contributing to an agent harness, agent runtime, or developer platform for AI agents
- Experience with Model Context Protocol (MCP) servers and tool ecosystems
- Experience building LLM evaluation pipelines and grader systems (LangSmith, Braintrust, custom frameworks)
- Experience with vector databases (pgvector, OpenSearch, Pinecone, Weaviate) and hybrid search
- Experience with Pulumi for infrastructure as code
- Familiarity with AWS Bedrock, SageMaker, Glue, Redshift, Athena, or QuickSight
- Experience with Azure and cloud migrations from Azure to AWS
- Experience with observability platforms such as Datadog or AWS CloudWatch/X-Ray
- Understanding of the AECO industry or document-heavy enterprise workflows
- AWS certifications (Solutions Architect, Data Analytics, or Machine Learning)
- Bilingual in French and English
Technical Skills & Tools
Languages
- Python, TypeScript/Node.js, C#/.NET
Agentic AI
- LangGraph, LangChain, Claude Agent SDK
- Model Context Protocol (MCP)
- LangSmith and evaluation tooling
LLM Platforms
- AWS Bedrock
- Anthropic Claude
- OpenAI
AI Coding Agents
- Claude Code, Cursor, GitHub Copilot
Data & Streaming
- Kafka/MSK, Kinesis, EventBridge
- Glue, Redshift, Athena, S3
Search & Vectors
- OpenSearch, pgvector, Pinecone, Weaviate
AWS Compute
- Lambda, ECS/Fargate, EKS, Step Functions
Infrastructure & Delivery
- Pulumi, GitHub Actions, Docker, Kubernetes
Observability
- Datadog, CloudWatch, X-Ray, OpenTelemetry
What We Offer
- A senior role with real influence over the platform that powers Newforma's AI strategy
- The chance to help build an agentic software factory, one of the most interesting problems in software engineering today
- Production-scale problems: billions of documents and emails, hundreds of thousands of users
- A strong AWS partnership with access to the latest AI and data services
- A collaborative Platform Engineering team working alongside our Lead Architect, DevOps, DevSecOps, and data engineers
- Competitive compensation and a comprehensive benefits package
- Professional development budget for conferences, courses, and certifications
- Flexible work environment with up to 10% travel for company and team events
Our Culture
At Newforma, We Value Team Members Who
- Are curious about our product and take initiative when problem solving
- Stay up to date with the latest technologies and techniques, especially in agentic AI
- Are open-minded and welcome feedback
- Want to be challenged as well as challenge others
- Are result-driven and don't hesitate to think outside the box to find solutions
- Are team players who respect colleagues and their opinions
- Follow best practices and clean code principles, whether the code is written by a human or an agent
Why Join Newforma?
Few engineers get to work on the data platform, the agent runtime, and the way software itself gets built, all at once. At Newforma, the Staff Platform Developer does exactly that. You'll help turn agentic AI from promising prototypes into dependable infrastructure, for our customers and for our own engineering teams.
If you love building platforms other developers rely on, you're fascinated by what makes AI agents actually work in production, and you want to help define how software gets built in the agentic era, we'd love to talk to you.
