CHABER LABSLLC

Rhode Island engineering company

Systems you can verify,
from the sensor to the decision.

Chaber Labs builds trustworthy AI/ML and embedded electronic systems. When one of our systems gives an answer or drives a control loop, you should be able to trust that result and check how it was reached.

Start a conversation Download capability statement (PDF)

Core competencies

Trustworthy AI / ML

  • Citation-faithful, hallucination-resistant retrieval-augmented generation (RAG)
  • Explainable AI (XAI), evaluation harnesses, and correctness red-teaming
  • On-prem and air-gap-friendly open-weight LLM deployment, plus QLoRA fine-tuning
  • Production generative-AI agents (tool-use / MCP) and SQL-over-data-lake analytics
  • Data engineering on AWS: ETL and streaming pipelines, a data lakehouse and warehousing, cataloging and governance, infrastructure-as-code, secure multi-account

Embedded & Electronic Systems

  • Electrical engineering: schematic capture, PCB design and prototyping, mixed-signal
  • Semiconductor and ATE test engineering, device characterization, bench instrumentation
  • Embedded firmware: ARM in C, bare-metal and Assembly (x86, ARM, AVR), real-time control
  • Power electronics: GaN power stages and grid-forming control loops
  • Deterministic networking and timing: custom L2 IEEE-1588 PTP, CAN, Modbus
  • Edge-to-cloud IoT, from device to firmware to secure connectivity to the cloud

How we work

The work runs from the physical world to the decision: we instrument the sensor, run it on real-time embedded firmware with precise timing, move the data securely across edge and cloud, and turn it into AI and analytics that hold up to scrutiny.

Full-stack in one engineer

From board bring-up and firmware through cloud to trustworthy AI. That range is unusual to find in a single contractor.

Verification first

Citation-faithful AI and test-driven hardware. The aim is to show the evidence, not just the output.

Active R&D

Self-funded power-electronics R&D with device-level SPICE design, real-time grid-forming control firmware, and custom IEEE-1588 time-sync in progress. Honest work in progress that keeps the skills sharp.

Representative experience

More than ten years designing and building software and electronic systems, including:

  • Consumer-robotics IoT cloud on AWS at large scale (10 to 20 million concurrent connected devices) as development lead, under GDPR and CCPA.
  • Connected medical devices: BLE firmware and a mobile companion app, with HIPAA and PHI handling, including technical oversight of an outside EE contractor (schematic and BOM review).
  • Structural and seismic sensing: an AWS IoT backend for sensor-data collection.
  • Industrial test automation (independent consulting): CAN, Modbus, and Protocol Buffers driving a 2-axis gantry and a rotation table, with embedded ARM in C and EE prototyping.
  • Analog and mixed-signal semiconductor test engineering.
  • Current work: production data engineering and generative-AI analytics, from ETL and streaming pipelines and a cloud data lakehouse to an agentic tool-use layer over curated data for natural-language analytics.

Who we work with

Defense & government

Local to the Newport and Quonset defense-industrial base in Rhode Island, with capability aligned to Navy priorities in explainable AI, secure data pipelines, and precise timing. A detailed, buyer-specific capability statement is available on request.

Regulated & technical industries

Teams in law, manufacturing, medical, and IoT who need answers that are auditable, not just plausible, and hardware that is tested rather than assumed.

Let's talk

Tell me what you are trying to verify, automate, or build. I reply personally.