Client work and our own projects are the two wheels that drive enginee’s engineering. In client work, our team embeds itself inside the systems and communities where the work happens, drawing on practical and academic expertise in machine learning (including generative AI), mathematical optimization, multimedia processing and HCI to build, together with our clients, systems that people actually use. Our own projects start from the ideas and concerns our members and collaborators freely pursue, and range from rapid app development to large-scale cultural projects, carried out through both research and practice.
case 01
Forecasting and visualizing future process loads,designing a process that continuously optimizes staffing and job sequencing
Everyone on the team, engineers included, went onto the factory floor. Starting from the gap between the process master data and a reality that keeps changing, we laid out the problems to solve on the way to optimizing the production sequence. We then developed a machine-learning feature that forecasts process loads, to build and update the process master. Used on the floor to see upcoming workloads, it grows more accurate with use, paving the way to a complete process master and an optimized production sequence.
case 02
Proper nouns, industry terminology:voice input that understands the local context of sound and meaning
By detecting at the audio layer whether, and where, registered words occur, and combining this with standard STT (speech-to-text), we built voice input that handles proper nouns and industry terms, a challenge particular to voice input in Japanese. We are rolling it out step by step as a voice-input app that runs entirely on the device, without sending audio over the internet (released on iOS only, as of November 2026).
case 03
Customer support automation with AI agents,supported end to end, from PoC to production
Client sector: E-commerce
#AIAgents #OperationsAnalysis #CustomerSupport
We joined a project applying AI agents to customer support, providing PM support, operations analysis and development. Our engineers stationed themselves at the call center and analyzed recordings of the work one by one to estimate the expected impact. When the ROI came into question, we proposed to the client’s executives that the project be halted. In the end, the project continued with its sights set on outcomes beyond efficiency, and we supported it through PoC, MVP development, production rollout and operations.
case 04
Supporting KPIs that balance medical safety and economics,and building a generative-AI platform for testing hypotheses
We created KPIs for each department, along with an investment decision workflow, that balance maintaining medical safety with economic viability, so that care stays stable and the hospitals can grow. We also built a dashboard with generative AI built in, so that busy front-line teams can keep running their own hypothesis testing.
case 05
Tracing the vast chain of translation
Technical partner: Media of Langue
“That word was translated into this word. And this word—” A non-profit project that links together the records of translation accumulated around the world, unearths the vast network that unfolds from them, and opens it up as a dictionary and map that anyone can explore. enginee contributes to Media of Langue as a technical partner, through data design and the implementation of its interactions.