WHAT YOU OWN

Everything the detection stack can hear is bounded by what the antenna captures, what the front end preserves, and what the SDR hands to the software as samples. You own that whole chain, from the electromagnetic model of the antenna to the demodulated symbol, and you own the measurements that prove the model was right. You report to the CTO.

• SDR signal characterization. Raw IQ from UAS control links and video downlinks, on USRP and bladeRF class hardware. Modulation, bandwidth, symbol rate, hopping pattern, framing. You build the GNU Radio and Python flowgraphs that capture, decimate, demodulate, and label. You tell the detection team what is actually in the air and you give them the reference signals to train against.
• Antenna and array modeling. Our antennas are designed in house. You model them before they exist: element design, impedance, pattern, gain, and polarization in HFSS, CST, FEKO, or openEMS. Direction-finding pairs and arrays: element spacing, mutual coupling, phase-center stability, and the bearing error each one produces. The simulation is your hypothesis; the VNA and the range are the test.
• RF front-end and link modeling. Link budgets from the airframe to the ADC. Noise figure, sensitivity, dynamic range, filter and LNA placement, intermodulation and blocker survival, cable and connector loss. Propagation over wooded, urban, and open terrain in Puerto Rico. You set the numbers the system is designed to.
• Fabrication and measurement. You build what you modeled and you measure it: RF PCB, printed and sheet-metal structures, phase-matched cabling. S11, return loss, bandwidth on the VNA. Gain and pattern in a chamber or by open-field comparison. Then you close the loop: where the measurement disagrees with the model, you find out why and you fix the model.
• Test bench and procedure. Instrumentation, calibration, controlled signal injection, repeatable written procedure. If a result cannot be reproduced next month, it is not a result.
• Over-the-air collection and ground truth. Live Group 1-3 targets, fixed and moving, multiple airframes including modified ones. Labeling protocol, versioning, metadata. The training corpus is a company asset and its quality is your responsibility.
• Government criteria characterization. Detect, track, and identify performance measured against published counter-small-UAS characterization criteria, documented cleanly enough to drop into a federal proposal without rewriting.
• Radome and enclosure effects. Detuning, loss, and pattern distortion from the housing, modeled first and measured second, with the mechanical engineer, so the antenna that passed on the bench still passes inside the box in Caribbean heat and humidity.

The model you build is the model the product is designed to. The number you measure is the number in the test report. Nobody downstream second-guesses it. That is the weight of the seat.

HOW WE WORK

We are small, and we move at a pace that only a small team can. Engineering runs on two-week cycles against a single shared plan. Every seat, RF, mechanical, electrical, and software, owns a slice of the same rig and has to land its objectives on the same date the others do. An antenna that is finished three weeks before the enclosure is not finished. A mount that is finished after the field test is a failure.

You scope your own work, commit to dates, say early and loudly when a date is at risk, and integrate with the rest of the stack instead of handing over a part and walking away. Pace is a requirement of the seat, not a culture slogan, and it is the reason a company our size can put fielded hardware in front of a government customer.

WHAT WE REQUIRE

B.S. in Electrical Engineering or Physics; M.S. in RF, electromagnetics, or communications is a plus and counts as experience. Roughly one to four years. Hands-on work counts wherever you did it: thesis, senior design, student competition, a personal project, or a bench in your apartment. We care that you have done it, not who paid you to.

• You have modeled something and then measured it. An antenna, an array, a receiver chain, or a link, simulated first and then built and characterized. Tell us where the model and the measurement disagreed and what you learned.
• Electromagnetics and antenna theory, applied. Radiation mechanisms, impedance and matching, gain, directivity, polarization, pattern, bandwidth. Dipole, monopole, patch and microstrip, Yagi, log-periodic, spiral, and arrays, and when to use each. Array factor, element spacing, grating lobes, mutual coupling, phase-comparison direction finding.
• EM simulation, hands-on. Working fluency in at least one of HFSS, CST, FEKO, or openEMS, including meshing judgment, port and boundary setup, and knowing when the solver is lying to you.
• SDR and digital communications. FSK and GFSK, BPSK, QPSK, QAM, OFDM, DSSS, and frequency-hopping spread spectrum, and how each looks in IQ. Sampling, decimation and filtering, symbol timing and carrier recovery, constellation and spectrogram interpretation. You can look at raw IQ and say what the signal is doing.
• SDR platforms, hands-on. GNU Radio, UHD or SoapySDR, Python with NumPy and SciPy for capture, DSP, and plotting. Real SDR hardware, including USRP or bladeRF class devices. Linux comfort.
• RF systems fundamentals. dB math, link budgets, free-space and terrain path loss, multipath and fading, noise figure, sensitivity, dynamic range, filters, LNAs, mixers, transmission line and connector loss. You reason in RF; you do not look it up mid-conversation.
• Lab instrumentation. Spectrum analyzer, VNA, signal generator, oscilloscope, power meter. You know what a calibration is for and why it expires.
• Prototyping. Comfortable at a bench with a soldering iron. Any of: RF PCB layout, sheet metal, 3D printing. You can get from a model to a physical part you can measure.
• Method over conclusions. If the measurement says the product underperforms, we need to hear it from you first, and we will act on it, not argue with it.
• U.S. person under ITAR.
• Onsite in Hatillo, Puerto Rico, and in the field, outdoors, carrying gear.

Preferred: MATLAB or Python channel and receiver modeling. Direction-finding algorithms (phase interferometry, MUSIC, correlative). RF PCB layout and 50-ohm design practice. Circular polarization. UAS or drone hands-on time. Clearance eligibility. Spanish.

WHAT WE OFFER

We pay a Puerto Rico competitive base and we win on everything around it: ownership of your domain, early equity, and coverage most companies our size do not offer.

• Base salary within the posted range, set by demonstrated capability, not by years on a resume.
• Early-employee equity. Stock options in iTerra Solutions, Inc., the Delaware parent, granted at pre-seed valuation. You join before the first priced round, which means you own a piece of the company at the lowest price it will ever be. Subject to board approval and standard vesting.
• Medical, dental, and vision, employee coverage fully paid by the company.
• Wellness benefit. A monthly stipend toward a gym membership.
• Flexible PTO.
• Direct access to the CTO and the founder. No layers. Your work is reviewed by the people who will ship it.
• Your work goes to the field in year one. Whatever you own will be characterized against a real government criteria set within twelve months. Very few engineers get that in their first five years.