Shota Iwami
Autonomous Driving / Platform Engineer
Vehicle Control · Sensors / SRE · Observability
Datadog Ambassador 2026Summary
Joined the autonomous driving team as a founding member and worked across the vehicle side, on both hardware and software. Now lead the vehicle team and own trajectory tracking control for the trajectories output by the E2E driving model.
Before that, worked on software platforms, SRE, and observability, building and operating backends for large-scale production services and their monitoring infrastructure. At newmo, first worked as a platform engineer for the taxi and rideshare dispatch services, building feature flags, CI/CD, and distributed tracing. Observability remains a core strength; selected as a Datadog Ambassador 2026.
Core Skills
- Vehicle Control: MPC (incl. contouring formulation / MPCC), PID, trajectory tracking, system identification (frequency response), closed-loop simulation, drive-by-wire
- Sensor Integration: Camera, LiDAR, GNSS/INS, CAN bus
- On-vehicle Edge Inference: NVIDIA DRIVE Orin (DRIVE OS)
- Robotics Visualization: Foxglove
- Hardware: sensor mount design / fabrication / wiring, CAD (Fusion360 / SOLIDWORKS)
- Languages: Go, C++, Python
- Backend & Data: gRPC / Protocol Buffers, Aurora / Spanner / DynamoDB
- Cloud & Infrastructure: Google Cloud / AWS, Terraform, CI/CD (GitHub Actions)
- Observability & SRE: OpenTelemetry, Datadog, load testing (k6)
Work Experience
newmo, Inc. - Lead Robotics Engineer / Autonomous Driving Software Engineer
April 2025 - Present
Autonomous Driving Development Office
Founding member of the autonomous driving team. From the start, worked across the vehicle side: bringing the autonomous driving software onto real vehicles, plus sensors, on-vehicle compute, and drive-by-wire. Now lead the vehicle team and primarily own trajectory tracking control. Also set the team's goals and coordinate the model-to-control interface with the planning team.
Vehicle Control (current focus)
- Designed and implemented an MPC-based trajectory tracker for the E2E model output, with actuator delay compensation, and drove a real vehicle autonomously with it
- To evaluate trajectory tracking independently of the E2E model, built a system that replays recorded human driving as timed reference trajectories, so different trackers can be compared on the real vehicle
- Built a closed-loop simulator that combines the actual driving model with identified vehicle dynamics, used to check controller changes before real-vehicle tests
- As team lead, planned the bring-up of a newly introduced development vehicle and owned the control side through its OEM-provided vehicle interface, from system identification to autonomous driving
Vehicle Platform Bring-up (initial focus)
- Converted an existing vehicle to drive-by-wire with our own system: microcontroller-based steering / braking / throttle control and the low-level control loop, later moving to control over CAN based on signal analysis
- Single-handedly brought the autonomous driving software onto that vehicle and designed, fabricated, and wired all of its sensor mounts, and reached closed-course autonomous driving with it
- Built an edge ML inference environment with ONNX Runtime on NVIDIA DRIVE Orin / DRIVE OS and integrated a GNSS/INS positioning unit into the stack
Vehicle Software Delivery
- Designed and built an OTA platform for the on-vehicle computers: OS images are signed and published from CI as OCI Artifacts, replicated to per-depot registries, and verified offline on the vehicle; in use for flashing the on-vehicle ECUs of the development vehicles
Technology Stack:
- Language: C++, Python
- Edge: NVIDIA DRIVE Orin (DRIVE OS)
- Sensing: LiDAR, GNSS/INS, Camera, CAN bus
- Control: MPC, system identification
- Platform & Tools: OCI Artifact (OTA), Foxglove, CAD
newmo, Inc. - Platform Engineer (Taxi / Rideshare)
November 2024 - March 2025
Drove trunk-based development by introducing feature flags (OpenFeature) and streamlining CI/CD. Built distributed-tracing observability with OpenTelemetry across services.
Error & Alert Operations Automation
- Designed monorepo error triage (automated runbook generation and optimized team-targeted alert routing); presented this work at Datadog DASH 2025
Technology Stack:
- Language: Go
- Infrastructure: Google Cloud (Cloud Run, AlloyDB), Envoy, Terraform
- Observability / CI/CD: OpenTelemetry, OpenFeature, Datadog, GitHub Actions
CyberAgent, Inc. - Backend Engineer
April 2022 - October 2024
AI Business Division, Co-creation Retail Media Division
Built and operated backends (Go) for large-scale production services on Google Cloud and AWS. In the internal "Sensing Seminar" program, also took part in research on in-store positioning using UWB and computer vision together with members of AI Lab.
New app for a major drugstore chain (April 2023 - October 2024)
Built the backend APIs, external service integrations, and observability infrastructure for an app with millions of active users. Drove trunk-based development to improve delivery speed.
Technology Stack:
- Language: Go
- Infrastructure: AWS (ECS on Fargate, Aurora Serverless v2, DynamoDB)
- Comms & Platform: gRPC (grpc-web / Envoy), Terraform, GitHub Actions, Datadog
CRM build & data migration for a foreign apparel company (April 2022 - March 2023)
Built a new CRM from scratch, integrating multiple brands and platforms. Responsible for member-facing features, external service integrations, data migration, and load testing.
Technology Stack:
- Language: Go
- Infrastructure: Google Cloud (Cloud Run, Cloud Spanner, Cloud Functions, Pub/Sub)
- Comms & Platform: gRPC / REST (grpc-gateway), Terraform, GitHub Actions, Datadog
GREE, Inc. - Full Stack Engineer (Internship)
January 2021 - September 2021
Aumo Inc.
Implemented backend infrastructure and tuned frontend performance for the media platform "aumo."
Technology Stack:
- Language: Ruby (Ruby on Rails)
- Infrastructure: AWS
hama, Inc. (formerly Space Entertainment Laboratory) - Airframe & Hardware Engineer (Internship)
January 2019 - March 2022
Joined the development of HAMADORI (then "UFB"), a seaplane-type fixed-wing UAV, from an early stage before it could even take off from water. Repeatedly worked on airframe design (e.g., motor mounts), fabrication, and on-site experiments in Fukushima, staying through to the point where it could take off from and land on ponds and the open sea and fly autonomously. Also took part in a stratospheric high-altitude balloon project for Nissin Yakisoba U.F.O. (a Guinness World Record for the highest-altitude smartphone live stream) and, with NS Solutions, a proof-of-concept for "turning the sky into a medium" by towing flags with drones.
Technology Stack:
- Airframe / Flight Controller: PX4 Autopilot, Pixhawk, Arduino
- Design & Fabrication: SOLIDWORKS (CAD)
Education
Tokyo Metropolitan University Graduate School
2020 - 2022
Master of Engineering, Aerospace Engineering
Thesis: "Flight Experiment of Four-dimensional Navigation for Drones Assuming Unmanned Aircraft Traffic Management"
Tokyo Metropolitan University
April 2016 - March 2020
Bachelor of Engineering, Aerospace Engineering
Publications
Opt-in Camera: Person Identification in Video Based on UWB-device Positioning, and Its Application to Opt-in Systems
MIRU 2024 (27th Meeting on Image Recognition and Understanding) / Demo Award (co-authored)
Software Design, May 2025 — Lead feature "Observability Frontiers: Fundamentals and an Introduction to OpenTelemetry"
Gijutsu-Hyoron (Technical Magazine)
Recognition
Datadog Ambassador 2026
Datadog's global community of technical experts and community leaders
Selected by nomination and panel review for advancing observability through technical talks, writing, and community engagement. First selected in 2026; membership is reviewed annually.
15th All-Japan Student Indoor Flying Robot Contest — Autonomous Flight Division: 5th place / 3D EXPERIENCE (CATIA) Award
Hosted by the Japan Society for Aeronautical and Space Sciences
Responsible for airframe design and fabrication (CAD: CATIA / Fusion360).
Conference Talks
CloudNative Days Winter 2026 (November 2026, accepted)
Topic: "Building an OTA Platform for Autonomous Taxis with OCI Artifacts"
Datadog DASH 2025 — New York, USA (June 2025)
Speaker (Datadog's flagship annual conference)
Topic: "Monorepo Error Management: Automated Runbooks and Team-Targeted Alert Distribution"
CloudNative Days SUMMER 2025 (May 2025)
Datadog Sponsor Session
Topic: "Building Observability Infrastructure from the Ground Up at an Early-Stage Startup"
Datadog Live Tokyo 2024 Reprise (December 2024)
Panelist (Datadog's official regional event)
Panel: "SREs Talk! Building Observability with Datadog — Challenges and Outlook from Product Companies"
CloudNative Days SUMMER 2024 (June 2024)
Topic: "Declarative Aggregation Management of Feature Flags Using OpenFeature and Code Generation"
Go Conference 2024 (June 2024)
Topic: "Building Go AST from Unified Diff Format Differences for Automatic Feature Flag Instrumentation"
DevOpsDays TOKYO 2024 (April 2024)
Topic: "Unified Management of Errors, Alerts, and Runbooks Using Code Generation to Reduce Operational Costs"
Go Conference mini 2023 Winter IN KYOTO (December 2023)
Topic: "Efficient Load Testing Strategies Using Go's Protoc Plugin"
Community Talks
Japan Datadog User Group Meetup #12 (August 2025)
Topic: "Monorepo Error Management: Automated Runbooks and Team-Targeted Alert Routing"
Jagu'e'r Observability SIG Meetup #3: Software Design (July 2025)
Topic: "Mastering OpenTelemetry Logs"
ROSCAFE: Visualizing and Optimizing Systems with Datadog (February 2025)
Topic: "Monorepo Error Triage: Team-Targeted Routing with Datadog"
MagicPod: Top 4 Load Testing Tools! User Insights into Effective Performance Testing Practices (September 2024)
Topic: "Building a Scalable and Reproducible Load Testing Platform with k6"
Findy: Learn from Real-World Use Cases! The Unknown World of Datadog (August 2024)
Topic: "Advanced techniques for leveraging Datadog's features"
OpenTelemetry Meetup #3 (June 2024)
Topic: "Feature Flags and OpenTelemetry"
Certifications
- Datadog Fundamentals (July 2023) - Datadog
- Professional Cloud Architect (April 2023) - Google Cloud