Consultancy for software development and research-based delivery — edge AI, autonomy, and systems that perform outside the lab.
MODE
FIELD-READY
STACK
EDGE · RESEARCH
LAT
OPTIMIZED
STATUS
AVAILABLE
Services
Software that ships and research you can trust — engineered for latency, reliability, and real operating conditions.
Production-grade systems across embedded, backend, and analytics — built for latency, reliability, and real operating conditions.
Research that leaves the lab. We turn papers, prototypes, and experiments into measurable, deployable outcomes.
End-to-end autonomy stacks for UAV, robotics, and edge AI — from sensing and inference to control and monitoring.
Engagement quiz
Four quick questions. Get a recommended engagement path — then book a call or send a brief.
Question 1 / 4
Resources & micro-offers
Free checklist library, or a fixed-price async audit. Upload notes here, then pay and submit your brief on the next page.
Open the library for Edge AI & UAV, ROS 2 / PX4, Jetson inference, BMS / EV, grid theft detection, Power BI KPIs, and more. Tick items in-browser — print or save as PDF.
Send your architecture notes or repo overview. Within 5 business days you’ll get a written audit: risks, latency bottlenecks, and a 30-day action plan for edge AI / UAV stacks.
Start engagement
Domains
Specialty zones across autonomy, perception, energy, and industrial intelligence.
Approach
A signal path from discovery to validation — research and software moving together.
Clarify constraints, success metrics, and the environment where the system must actually work.
Architect software and research paths together — simulation, models, and interfaces before code scales.
Implement with production discipline: measurable latency, fault handling, and field-ready interfaces.
Benchmark, document, and deliver — including research outputs your stakeholders can act on.
Selected work
Representative engagements spanning autonomy, edge AI, energy systems, and research publications.
Production-oriented edge stack: real-time inference, MQTT telemetry, and PX4 waypoint injection for disaster response.
Robust offboard communication-loss handling and MAVLink timeout strategy validated for flood-detection missions.
Cost-effective deep learning and time-series forecasting for utility grid theft detection — IEEE publication.
Full vehicle and battery models: regenerative braking, cell balancing, and drive-cycle optimization in Simulink.
Book a call
Pick a slot to discuss scope, constraints, and whether a pilot or async audit is the right next step.
About
Software engineered for the field. Research that ships — led from Taiwan, India, and remote.
Embedded Systems · Edge AI
Embedded software engineer focused on production-grade autonomous systems — edge AI inference, real-time UAV pipelines, and industrial data platforms. Integrating hardware, firmware, and ML into deployable systems with measurable latency and reliability outcomes. Edgefield exists to close the gap between lab research and field reality: helping teams ship autonomy and edge AI that holds up outside the demo — with clear engineering judgment, not slideware.
View personal portfolioInfrastructure · Field Operations
Proactive and adaptable engineer educated in India and Nepal, with proven expertise in managing urban and district infrastructure projects. Known for applying innovative problem-solving techniques to maintain high engineering standards and safety while effectively reducing project delays and improving operational efficiencies in diverse environments.
Contact
Register and sign in to request consultancy or product development. Share your brief — we’ll reply from aykumar2106@gmail.com.
aykumar2106@gmail.com