About LMAi Labs
Software built by someone who ran these programs inside an agency
Small and mid-size transit agencies sit on real data — GTFS, ridership, Census ACS — but the vendors selling planning software are priced and staffed for the largest metros. LMAi Labs exists to close that gap.
Founder
Baofeng Dong
I’m Baofeng Dong, a tech program manager and tech leader with 10+ years leading cross-functional teams in deploying new products and processes end-to-end and building new applications — across Amazon, Confluent, and public transit.
At Confluent, I lead Cloud Platform Control Plane programs powering real-time AI and agentic workloads, and built an AI-powered platform discovery tool (FastAPI + React + Claude API) that consolidated 6,800+ knowledge artifacts and reduced platform research from days to minutes for 100+ engineers. At Amazon, I led engineering, product, operations, and process teams in deploying a new generation of warehousing tools to millions of associates at thousands of warehouses worldwide — resulting in a 31% reduction in scanning errors and $20MM in annual transportation savings, contributing to over $600M in cost savings across 700+ fulfillment sites.
Before Amazon, I spent years driving innovation tech at a public transit agency — and that’s the work that shaped why LMAi Labs exists. I pioneered and built TriMet’s tablet-based on-board rider survey program, replacing the paper-based survey with an Open Data Kit + in-house Android stack. The result: ~30% cost savings versus paper, materially better data quality, and a 92% reduction in lead time for survey results. I also deployed Fleetwatch and other enterprise IT solutions, and saw firsthand that small/mid-size transit agencies don’t lack data — they lack the IT capacity to tap into it for service planning and Title VI equity analysis.
Outside client work, I serve as a panel member on multiple FTA- and APTA-sponsored Transit Cooperative Research Program (TCRP) projects, guiding consultants’ work, and I’m a frequent presenter at local and national conferences on new-technology implementation and project management. LMAi Labs is how I bring all of this together — practical software for small/mid transit agencies that don’t have massive engineering teams.
Mission
Bridge the gap between the data agencies already have and the analysis they can run in-house
Small and mid-size transit agencies sit on rich data — GTFS, ridership records, Census ACS demographics — but rarely have the IT capacity to put it to work for service planning and Title VI equity analysis. LMAi Labs bridges that gap: flat-priced, self-serve software shipped by someone who’s actually run these programs inside an agency.
What we build
A product, a methodology, and select custom work
The product
Transit Planner
Cost a service change, run Title VI equity analysis, and generate board-ready reports — self-serve, on your own GTFS and Census data.
See the product →Methodology
FTA Title VI & compliance work
The four-fifths disparate-impact test, major-change triggers, and a published methodology aligned to FTA Circular 4702.1B.
Read the methodology →Custom work
Custom data & AI tools for public agencies
Need something Transit Planner doesn't do yet? We take on select custom data and AI projects for public agencies.
Talk to us →Run your first scenario this afternoon
Create a free workspace, upload your GTFS, and see your network costed in the browser. No setup call, no credit card. You are in your own workspace the same afternoon.