At Kandou, we are redefining the economics of AI infrastructure. Our mission is to democratise AI by significantly reducing the Total Cost of Ownership (TCO) of hardware systems — a critical barrier to scalable adoption.
Our proprietary MIMO-over-copper technology powers a high-performance, chiplet-based AI memory fabric that is both scalable and energy-efficient. Unlike traditional interconnects, our solution reduces power consumption significantly while preserving high bandwidth and ultra-low latency — unlocking unprecedented efficiency for AI training and inference at scale.
Kandou’s architecture is not just an incremental improvement — it’s a foundational shift in how AI hardware is built for the future.
We are actively seeking an Embedded Software Engineer based in Hyderabad:
Experience
7+ years' experience in designing, coding and debugging real-time applications in C and/or C++ in an embedded environment
Experience in porting and use of RTOS: Zephyr(preferred), FreeRTOS etc.
Knowledge of RISC-V CPU architecture; experience writing bare-metal or RTOS-based firmware targeting RISC-V cores (RV32/RV64)
Board and SoC bring-up with JTAG and other debug mechanisms
FW Update, bootloaders, secure boot
Debugging at application, driver and hardware levels
Interfacing to hardware and peripherals: UART, SPI, I2C/SMBus, I3C
Commitment to use of modern SW development and test workflow tools (CI / GitLab / Makefile / CMake / TDD etc.)
Knowledge of management and sideband protocols including SMBus, I3C, MCTP, CMIS
Skills
Working with Heterogeneous multicore systems, including both scalar and vector processors
Real-world experience of high-speed signalling technologies: PCIe, CXL — particularly link training, equalization and retimer/redriver interaction with host and device PHYs
Firmware development for Active Electrical Cable (AEC) or retimer/redriver ASICs: cable diagnostics, eye margin monitoring, temperature/voltage telemetry and power-state management
Memory and network interfacing, virtual memory, simulation and co-simulation
Firmware Update approaches and algorithms
Use of simulation and/or FPGAs for pre-silicon development
Responsibility
Develop, test and maintain Embedded Software for Kandou AI’s new ASIC products
Real-time Firmware design on high-speed heterogeneous SoC platform
Work with Architects and Lead engineers to design and implement according to system and module specifications
Software design and support for our silicon design infrastructure and workflow
Participate in and promote current best practices in test, review, integration, automation and delivery of quality software products
Visit: www.kandou.ai/careers
Skills Required
- 7+ years of experience designing, coding, and debugging real-time embedded applications in C and/or C++
- Experience porting and using an RTOS such as Zephyr or FreeRTOS
- Knowledge of RISC-V CPU architecture and experience developing bare-metal or RTOS-based firmware for RV32/RV64 cores
- Experience with board and SoC bring-up using JTAG and other debugging mechanisms
- Experience with firmware updates, bootloaders, and secure boot
- Ability to debug at application, driver, and hardware levels
- Experience interfacing with UART, SPI, I2C/SMBus, and I3C peripherals
- Experience with CI, GitLab, Makefile, CMake, TDD, and modern software development workflows
- Knowledge of SMBus, I3C, MCTP, and CMIS management and sideband protocols
- Experience with heterogeneous multicore systems using scalar and vector processors
- Experience with PCIe and CXL high-speed signaling, including link training, equalization, and retimer/redriver interaction
- Firmware development experience for active electrical cables or retimer/redriver ASICs
- Experience with memory and network interfacing, virtual memory, simulation, and co-simulation
- Experience using simulation and/or FPGAs for pre-silicon development
What We Do
Kandou AI is a fabless semiconductor company developing high-speed connectivity for AI infrastructure. Its patented Copper MIMO, also known as Chord Signaling, sends correlated signals at data-processing speeds of 448 Gbps and beyond, providing an alternative to optical interconnects. The company’s technology targets AI data-movement bottlenecks and memory-wall constraints, while its platforms emphasize improved efficiency, lower power and cost, and scalability.
