Senior VoIP / SIP Engineer — Kamailio + RTPengine + Asterisk SBC Build

# Senior VoIP / SIP Engineer — Kamailio + RTPengine + Asterisk SBC Build (3 months, near full-time) ## The role We are building out a production carrier-grade voice estate and need an experienced VoIP engineer to lead the telephony infrastructure work end to end. This is hands-on build work, not advisory. You will design, stand up, harden and document a SIP edge and media layer that terminates PSTN traffic from US carriers and routes it into application logic behind it. You will be the telecom specialist on the project. Our in-house team covers application development, AI/ML, cloud infrastructure and delivery — we need the person who owns the SIP layer and knows why calls actually break. ## Must-have experience You should have built and run this stack in production, not just read about it: **SIP edge / SBC** - Kamailio (or OpenSIPS) as a production SBC — routing logic, `dispatcher`, `dialog`, `htable`, `pike` / `ratelimit`, `permissions`, topology hiding, TLS termination - KEMI scripting (Lua or Python) or strong native cfg scripting - RTPengine — media relay, kernel module, SRTP/DTLS, NAT traversal, in-kernel vs userspace forwarding trade-offs **Media / application layer** - Asterisk 20/22 — PJSIP stack, dialplan, ARI and AMI, queues, realtime architecture, ODBC/PostgreSQL backends - Codec handling and transcoding cost management (G.711, G.729, Opus) **PSTN and carrier integration** - Live SIP trunk integration with US carriers (Bandwidth, Telnyx, Inteliquent, Peerless, Twilio or equivalent) - E.164 normalisation and header normalisation across multiple carriers — ANI, P-Asserted-Identity, RPID, Diversion, `verstat` - DID inventory management, LERG/LRN awareness, porting workflows - STIR/SHAKEN — inbound verification and outbound attestation, `res_stir_shaken`, Identity header handling, certificate/CRL management **Operations** - High availability for a stateful SIP edge — active/active or VRRP, state replication, failover behaviour under load - Homer/HEP, sngrep, tcpdump — you debug at packet level as a matter of routine - SIPp load testing, CPS and concurrency capacity planning - Linux (Debian/Ubuntu), systemd, nftables/iptables, kernel tuning for high-PPS voice workloads - Infrastructure as code — Ansible and/or Terraform. We do not want snowflake servers. - Prometheus/Grafana instrumentation, CDR accuracy and audit logging ## Strongly preferred - Contact centre telephony — queueing, IVR, call recording architecture - Robocall/fraud mitigation at the SIP layer — reputation feeds, velocity controls, ANI validation, do-not-originate handling - US regulatory familiarity: FCC call blocking rules and SIP 603+, Robocall Mitigation Database, PCI DTMF suppression, E911 / Kari's Law / RAY BAUM'S Act - FreeSWITCH, WebRTC/SIP over WSS - Experience handing an estate over to an in-house team that will run it afterwards ## What we care about **Documentation and handover are part of the deliverable, not an afterthought.** Our team will operate this once you are done. Everything you build needs to be reproducible from code, with runbooks. If you have only ever configured boxes by hand, this is not a fit. We also want someone who says "that won't work" when it won't. You'll have direct access to the technical decision maker and we'd rather argue about the design early than discover the problem in production. ## Engagement - **Duration:** 3 months, with a strong chance of extension into an ongoing retainer - **Commitment:** 30–40 hrs/week — this is a near full-time engagement, not a few hours on the side - **Overlap:** minimum 4 hours daily overlap with UK business hours; some US (Central/Eastern) call availability - **Location:** fully remote, anywhere - **Rate:** competitive and negotiable for the right person — tell us your rate, don't guess ours - We work in two-week increments with defined outcomes. First increment is a paid scoped piece of work so we both find out quickly whether this works. Individual contractors preferred. If you're applying as an agency, name the specific engineer who will do the work and include their background. ## To apply Skip the template cover letter — we read the answers, not the boilerplate. Please answer: 1. Describe a production Kamailio (or OpenSIPS) + RTPengine + Asterisk estate you personally built. Topology, peak concurrent calls, carriers, and what broke in the first month. 2. How do you reject an inbound call at INVITE without establishing media, and which SIP response codes do you use in which circumstances — and why does the choice matter? 3. Two US carriers deliver the same call with different header treatment. Walk us through how you normalise ANI and caller ID authentication status so downstream logic sees one consistent shape. 4. What is your approach to high availability on the SIP edge specifically, and what state do you actually have to replicate? 5. A customer reports one-way audio on roughly 5% of calls. Talk us through your first hour. Short, specific answers beat long ones. A link to config you've written, a GitHub profile, or a description of a genuinely hard bug you fixed will get our attention faster than a CV.