Firmware Lead

Location
Hyderabad
Workplace
Hybrid
Employment
Full-time

About this role

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.

As Firmware Lead, you will lead implementation and delivery of real-time embedded firmware for Kandou's Randa 1.6T Active Electrical Cable platform. This is a coding-heavy technical leadership role: you will spend the majority of your time writing and reviewing C/C++, debugging firmware and hardware interactions, and supporting pre-silicon and post-silicon bring-up. You will turn the Firmware Architect's system design into production code, own implementation quality and delivery, and mentor engineers while remaining active in the codebase.

Open locations: India (Hyderabad or Bangalore), UK (Reading or Cambridge), Switzerland (Lausanne).

Responsibilities

• Own the software stack above the AEC retimer firmware on the 1.6T Active Electrical Cable platform: the CMIS-compliant host-management interface, test-orchestration software, the field-telemetry pipeline, the customer-facing SDK and tooling, and OSFP Multi-Source Agreement compliance. This role is customer-facing on software-integration questions for hyperscaler and neo-cloud segments.

• Architect the CMIS 5.x management stack (SFF-8024 identifier conventions, CMIS page layouts, OIF CMIS message flows); own the DataPath State Machine, Module State Machine, Application Selection logic, Versatile Diagnostic Monitoring (VDM) and failure-recovery transitions, and define byte-level vendor page assignments (B0h to FFh) for Kandou-specific extensions.

• Design the Customer Adaptation Layer (CAL) and its API boundary so per-customer integration is cheap by construction without forking the core codebase; own hook-point semantics, the Customer Integration Package (CIP) delivery model, and the CER extension-request process with SLA-bound versioning.

• Drive multi-vendor host interoperability validation across hyperscaler NIC and switch-ASIC platforms (Meta, Microsoft, Oracle, Google, AWS) plus host BMC implementations; own the interop test matrix, the failure-mode catalog, the regression suite, and represent Kandou software in customer technical escalations.

• Architect the manufacturing test-orchestration software above the firmware’s primitives: test sequencing, pass/fail decision logic, ATE integration, result logging and yield analytics; drive on-chip stimulus-response analysis (CTM/HTDR) without external pattern generators, oscilloscopes or TDR equipment.

• Architect the telemetry pipeline from firmware buffers through the CMIS transport into customer host consumption and, where policy permits, into Kandou cloud analytics for cross-fleet learning; own the data-format specification, compression and retention schema, privacy boundaries, and host-side APIs (REST, gRPC or customer-specified).

• Own the customer-facing SDK and tooling: management-library bindings (C, Python, and where required Go or Rust), diagnostic command-line tools, log-analysis utilities and reference integration examples; own the SDK release cadence, semantic versioning and backward-compatibility contract.

• DevOps and release engineering (key responsibility): own the build, test and release-engineering backbone for the full software platform — multi-target CI/CD pipelines, Kconfig-driven build variants and per-customer image selection (OPN suffix with fuse-based misdelivery protection), conformance and hardware-in-loop test automation, JSON-schema validation gates, artifact signing and software supply-chain integrity, infrastructure-as-code for interop and HIL labs, secrets management (e.g. HashiCorp Vault, AES-256), and NDA-isolation enforcement through repository ACLs and CI policy checks.

• Author and maintain normative software architecture specifications using SHALL/MUST/SHOULD language with REQ-XXX-YYY traceability; drive architecture-review cadence with the Firmware Architect, ASIC design, product management, manufacturing test engineering and field application engineering; contribute to OSFP MSA, OIF and CMIS working groups..

Essential Experience

• 8 to 10 years post-degree shipping high-speed interconnect software: module software, management stacks, test software or customer SDKs.

• Device-adjacent software experience on AEC, AOC, optical transceiver, retimer module or networking silicon; must not be pure host or cloud.

• Working implementation experience with CMIS and OSFP MSA (any recent revision), including conformance-test ownership.

• Management and sideband transport implementation: I2C/SPI, MCTP, SMBus, I3C.

• Hyperscaler or tier-one OEM interoperability experience across NIC platforms, retimers, switch-ASIC platforms and host BMC implementations.

• DevOps depth: CI/CD for multi-target builds, conformance-test-harness design, hardware-in-loop at the module level, build systems (Make/CMake, Kconfig), Git-based workflows and automated release pipelines.

• Strong C and Python, with commitment to modern development and test workflow (TDD, static analysis, automated regression, CI/GitLab).

• BS in Electrical Engineering, Computer Engineering or Computer Science (minimum); MS preferred.

Useful experience

• Platform-architecture approaches that isolate customer-specific logic behind a stable API (adaptation layers, plugin/hook models) to prevent fork-per-customer divergence across N codebases and N test matrices.

• Secure software supply chain: image signing, secure boot and root-of-trust (Caliptra or equivalent), A/B firmware-update orchestration and rollback.

• Infrastructure-as-code and lab automation: containerized hardware-in-loop, pytest-based DUT farms, OSFP host emulation and FPGA host adapters.

• Secrets and NDA isolation in practice: per-customer key management, AES-256 encryption, repository ACLs, audit trails and incident-response tooling.

• Manufacturing test economics and capital-equipment displacement (TDR, BERT, scope-based eye capture); appreciation of cryptographic algorithms and disciplines.

• Direct customer engagement with Meta, Microsoft, Oracle, Google or AWS, including qualification negotiation and field-issue triage with named engineering counterparts.

• Knowledge of SFF-8024 and SFF-8636 identifier conventions and legacy-page support; management-plane exposure of IEEE 802.3ck / 802.3dj link-layer behaviors (FEC telemetry, link-training diagnostics).

• Contributions to OSFP MSA, OIF or SFF working groups (proposals, reviews or chair roles).

• Exposure to safety-critical or regulated software (telecom, automotive, aerospace) that translates to hyperscaler change-management bars; published open-source contributions in the transceiver-management or module-software space

If this is the role you have been looking for and you want to be part of a growing Company, with an exciting future then we would love to hear from you. Together We Kandou It!!

Apply to Kandou AI

Clera fills in Kandou AI's application for you.

How it works

Open the role in Clera

Sign in, and this role opens on your Clera job board, ready to apply.

Clera fills in the application

Clera fills the form from your profile and resume, and asks what's missing.

Track it in Clera

Each application you send shows up in your Clera tracker, sorted by stage.

Want Clera to find roles for you?

Ask your AI about this role

It knows you. Clera gives it the role.

Does this role fit my profile?

First time? Connect Clera once, about a minute.
One click above pre-fills the connector. Or by hand: Settings > Connectors > Add custom connector, paste https://mcp.getclera.com. Either way, sign in with the account you use on Clera and approve the consent screen.
https://mcp.getclera.com

Sign in with your Clera account. MCP docs

About Kandou AI

Founded in 2011, Kandou is the innovative leader in high-speed, energy efficient, chip-to-chip link solutions critical to the evolution of the electronics industry. Kandou enables a better-connected world by offering disruptive technology through licensing and standard products that empower the devices we use every day to become smaller, more energy efficient and more cost effective.

Kandou has a strong IP portfolio that includes Chord™ signaling, which has been adopted into industry specifications by JEDEC and the OIF. These innovations and implementations deliver a fundamental advance in interconnect technology that lowers the power consumed and improves the performance of chip links, unlocking new capabilities for customer devices and systems.

Kandou is a fabless semiconductor company headquartered in Lausanne, Switzerland with offices in Europe, North America and Asia.

Team size
51-200 employees people
Founded
2011
Industry
Semiconductor Manufacturing
Workplace
Hybrid
Locations
Hyderabad, Telangana, India