Company brief · updated 2026-10-18
reflection.ai · software industry · Make intelligence open and accessible to all
Today: Our forward-deployed team works alongside your teams from ideation through application design, development, and deployment
Roles: Forward Deployed Engineer, Lead - AI Engineer · Forward Deployed Engineer - AI Engineer · Forward Deployed Engineer, Lead - LLM Post-training · Forward Deployed Engineer - LLM Post-training · Engagement Manager
Requirements mention: agent orchestration · RAG pipelines · fine-tuning · customer-facing environments · public cloud · VPC · on-premises · AI deployment lifecycle · Statements of Work
Inference: the company is building out a forward-deployed engineering function across New York, San Francisco and Seoul to adapt its open-weight models for specific enterprise customers. Seven of the 46 open roles sit in Commercial.
Today: Rigorously tested for safety and built for efficient performance
Roles: Member of Technical Staff - Evaluations · Member of Technical Staff - Research Software Engineer - Safety Evaluations Infrastructure · Member of Technical Staff - Pre-Training Evals Infra · Member of Technical Staff - Safety · Research Program Manager - Model Data & Evaluations Flywheel · Member of Technical Staff, Data Flywheel
Requirements mention: evaluation infrastructure · red-teaming and adversarial evaluation pipeline · sandboxed execution environments · audit logging · evaluation frameworks
Inference: the company is building evaluation and safety systems that feed into model release decisions, which two postings state directly.
Today: Infrastructure: open models served through an API platform or in customers' own environments
Roles: Member of Technical Staff - Data Platform · Member of Technical Staff - Engineering Lead, Data Platform · Member of Technical Staff - Shared Services · Member of Technical Staff - Engineering Lead, Shared Services · Member of Technical Staff - Platform Foundations · Member of Technical Staff - Engineering Lead, Platform Foundations · Member of Technical Staff - Applications · Member of Technical Staff - Engineering Lead, Applications
Requirements mention: production-grade pipelines at scale · Batch and streaming workloads · SLI/SLO design · Idempotency · backpressure · Terraform · Bazel · customer-facing surfaces
Inference: the company is building shared infrastructure for research, training and production, mostly internal but including customer-facing surfaces from the Applications team. Engineering has 10 open roles, second only to Research.
Today: Open-weight models, including Beam, a 501B open-weight model
Roles: Member of Technical Staff - Pre-Training Infra · Member of Technical Staff - Compute Platform · Member of Technical Staff - Engineering Lead, Compute Platform · Research Program Manager - Research Infrastructure · Member of Technical Staff - Research Software Engineer
Requirements mention: distributed training systems · NCCL · multi-cloud scheduling · fault tolerance · training throughput · Kubernetes-based platform
Inference: the company is scaling its training infrastructure on a Kubernetes-based platform that spans several neo-clouds, according to the Compute Platform postings.
Today: Open-weight models you can inspect, fine-tune, and deploy
Roles: Member of Technical Staff - Web Crawl Engineer · Member of Technical Staff - Data Ingestion Engineer · Member of Technical Staff - Multilingual Data · Member of Technical Staff - Data Quality Engineer (Pre-training) · Member of Technical Staff - Data Quality Engineer (Post-training) · Human Data Strategist
Requirements mention: URL frontier management · web-scale datasets · language identification · tokenizer design · LLM-as-a-Judge · expert data collection
Inference: the company is building its own pipeline for web-scale and human expert training data, with multilingual data a named focus.
Today: AI systems that give enterprises, governments, and sovereign entities control over their intelligence
Roles: MTS Lead, Identity and Access Management · MTS Lead, Product Security · MTS, Technology & Security Engineering · MTS Lead, Technology & Security Engineering · Compliance Manager · GRC Engineer · Sovereign Deal Lead
Requirements mention: zero-trust architectures · WebAuthn / FIDO2 · FedRAMP · CMMC · ITAR/EAR · FAR/DFARS · enterprise agreements
Inference: the security and compliance postings name US federal and defense frameworks alongside a sovereign deal role, which points to government customers being part of the plan. Sales capacity for these deals is one role so far.
| Department | Open |
|---|---|
| Research | 16 |
| Engineering | 9 |
| Security | 5 |
| Commercial | 5 |
| Technical Program Management | 3 |
| Finance | 2 |
| Legal | 2 |
| People | 1 |
| Sovereign | 1 |
| Other | 1 |
Research (15), Engineering (10) and Commercial (7) hold 32 of the 46 open roles, so hiring centers on model research and the platforms behind it, with a smaller customer-facing team; 11 roles opened in the last 30 days, against 5 in the prior 30.
Seniority: Mid 12, senior 8, manager 10, staff 3, principal 1, unknown 12.
Skills named most: Python (13), Kubernetes (12), software engineering (6), data pipelines (6), CI/CD (6), SOC 2 (5), ISO 27001 (5), Spark (5), Terraform (5), fine-tuning (5), APIs (4), Rust (3)
Locations: New York, NY (23), San Francisco, CA (21), Seoul (2). No remote roles are listed.
If you're applying: If you have Python, Typescript, RAG and agent orchestration experience, look at the Forward Deployed Engineer - AI Engineer roles in New York and Seoul. If you build data pipelines (Spark, Beam, Flink, Kafka) or web crawlers, look at the Data Platform and Research data roles. None of the roles are listed as remote, so check each posting's location before applying.
If you're selling or recruiting: Research (15) and Engineering (10) have the most open roles, and 11 roles opened in the last 30 days. Postings name the tools in use: Spark, Flink, Kafka, Iceberg, BigQuery and Snowflake for data; Terraform, Pulumi and Kubernetes for infrastructure; Vanta for compliance; Okta for identity; and Ashby for recruiting. The Compute Platform lead posting lists managing vendors, and the Sovereign Deal Lead role covers government and sovereign-backed deals.