AI Software Engineer

Arva Intelligence•Houston, TX

2d•Remote

About The Position

We are seeking an AI Software Engineer to design, build, and operate the AI-enabled services at the core of the Arva platform. This role sits at the intersection of applied AI engineering and production software development: you will build agent-based systems using frameworks such as LangGraph and LangChain, and you will also build the durable application and data infrastructure including APIs, pipelines, and interfaces that make those systems reliable at scale. This is a hands-on engineering role for someone who has moved past their first job, ships production code independently, and is looking to apply modern AI techniques to a domain where correctness genuinely matters.

Requirements

2–5 years of professional software development experience building and shipping production systems.
Strong proficiency in Python, including experience with modern application frameworks, testing, and packaging practices.
Demonstrated experience building applications or services that use large language models or other machine learning models in production.
Working knowledge of relational databases and SQL, with practical experience designing schemas and writing performant queries.
Experience developing and consuming RESTful APIs.
Familiarity with cloud infrastructure (AWS preferred) and standard software delivery practices including version control, CI/CD, and automated testing.
Ability to work independently on ambiguous problems, and to communicate technical trade-offs clearly to both technical and non-technical colleagues.
Bachelor's degree in Computer Science, Software Engineering, a related technical field, or equivalent practical experience.

Nice To Haves

AI frameworks — Building agentic and LLM applications with tools like Amazon Bedrock, LangGraph and LangChain.
Model architecture — Familiarity with Mixture-of-Experts (MoE) architectures, model routing, and ensemble or multi-model system design.
Backend — Production experience with Python and the Django framework.
Frontend — Proficiency with TypeScript, JavaScript, and Vue.js.
Data storage — Advanced PostgreSQL experience, including performance tuning and the PostGIS extension.
Geospatial — Experience with geospatial data formats, coordinate systems, raster and vector processing, satellite or remote sensing imagery, and libraries such as GDAL, rasterio, GeoPandas, or Shapely.
Data processing — Building large-scale ETL/ELT pipelines, batch and streaming workflows, and orchestration tooling.
AWS — Depth across services such as Lambda, S3, ECS/Fargate, RDS, Batch, SQS/SNS, Step Functions, and IAM.
Integrations — Designing and maintaining robust third-party and partner API integrations.
QA/QC — Establishing automated quality assurance and quality control processes for software, data, and model outputs.
Domain — Prior work in agriculture, agronomy, remote sensing, climate, energy, or environmental markets.

Responsibilities

Design, build, and maintain LLM-powered and agentic workflows using tools like LangGraph and LangChain, AWS Bedrock AgentCore, including tool use, retrieval, memory, orchestration, and multi-step reasoning pipelines.
Develop and productionize AI/ML services in Python, including work with Mixture-of-Experts (MoE) and other multi-model architectures for routing, specialization, and ensemble prediction.
Build and extend backend services and APIs in Django, with PostgreSQL (and PostGIS) as the primary data layer.
Implement front-end features and internal tooling in TypeScript and Vue.js to expose AI capabilities and analytics to users.
Build and optimize data processing pipelines that ingest, clean, normalize, and enrich large volumes of geospatial, agronomic, imagery, and third-party data.
Develop and maintain integrations with external partner and customer APIs, including authentication, rate limiting, schema mapping, error handling, and monitoring.
Deploy, scale, and operate services on AWS using managed compute, storage, database, queueing, and serverless offerings.
Design and execute QA/QC processes for both software and data — automated testing, validation rules, anomaly detection, model evaluation, and regression checks — so that model outputs meet the accuracy and auditability standards our customers and verification bodies require.
Participate in code review, architecture discussions, and technical planning; write clear documentation for the systems you build.
Partner with data scientists, agronomists, and product stakeholders to translate scientific and business requirements into production systems.