We build the network and the software that runs on it
Telecorenetwork is an enterprise cloud communications provider. We operate our own softswitch platform across 42 points of presence, hold our own carrier interconnects, and carry more than 2.4 billion minutes a year for 1,800+ enterprise customers. When something goes wrong on a call, there is nobody else for us to blame.
Why Telecorenetwork exists
The company started with a frustration that anyone who has run enterprise voice will recognise.
Buy from a legacy carrier and you get a solid network attached to software written in another decade: provisioning by email, CDRs by spreadsheet, a portal that logs you out every fifteen minutes. Buy from a modern software vendor and you get a pleasant portal and an API — sitting on top of wholesale minutes bought from someone else, three layers of resale away from the route your call actually takes. When quality drops on a Tuesday afternoon, the vendor opens a ticket with their supplier, who opens a ticket with theirs.
Our founders had spent years on both sides of that trade: two from a Tier-1 carrier's interconnect team, one from a contact centre software company that had been burned by exactly this chain. In 2016 they built a small softswitch cluster in Ashburn, signed their own interconnect agreements rather than reselling, and put an API in front of it on day one.
That is still the whole idea. We own the switching layer and the network layer, we publish the numbers those layers produce, and we staff the phone with the engineers who operate them. Everything else on this site is a consequence of that decision.
Talk to our teamTelecorenetwork at a glance
Regional offices in London, Frankfurt, Singapore and São Paulo. Network figures are drawn from the trailing twelve months and refreshed quarterly.
Ten years of building the platform
No acquisitions to integrate, no platform rewrites forced by a merger. The switch we run today is a direct descendant of the first cluster.
First softswitch cluster in Ashburn
Three founders, one rack in Northern Virginia and a single carrier interconnect. The first production customer was a 40-seat insurance brokerage replacing two PRIs. The provisioning API shipped in the same quarter as the portal, which set the pattern for everything since.
First European PoP and EU data residency
Frankfurt went live as PoP number seven, followed by London four months later. Customer recordings, CDRs and metadata could be pinned to EU storage from the day the region opened — a requirement we were asked for by a German logistics customer before GDPR enforcement began.
Wholesale interconnect programme
Remote-work demand pushed traffic past 400 million minutes a year. We opened direct interconnects with 38 national operators and launched the wholesale business, letting carriers and resellers buy the same routes our enterprise customers use rather than a separate B-stock.
CPaaS API and programmable voice launch
The internal control API became a public product: REST endpoints and webhooks for call control, conferencing, recording and number provisioning, with SDKs for Node, Python, Java, Go, PHP and .NET. SMS followed two quarters later on the same authentication and billing model.
SOC 2 Type II attestation and ISO 27001 alignment
Completed our first Type II observation window covering security, availability and confidentiality, and brought the ISMS in line with ISO/IEC 27001. HIPAA business associate agreements and PCI-DSS SAQ-D recording controls were productised in the same programme.
Two billionth minute and 42 PoPs
The platform crossed two billion carried minutes in a single year, the PoP count reached 42 across six regions, and the customer base passed 1,800 enterprises. Median edge latency held at 18ms and average MOS at 4.42 through the whole expansion.
# Carried minutes by year (billions) 2016 0.01 1 PoP 1 interconnect 2018 0.18 9 PoPs 7 interconnects 2020 0.41 17 PoPs 38 interconnects 2022 1.05 28 PoPs 64 interconnects 2023 1.62 34 PoPs 81 interconnects 2025 2.40 42 PoPs 110 interconnects # Held steady through every expansion availability 99.999% average MOS 4.42 edge latency 18 ms median P1 response 4 min median
Six operating principles
These are not values on a wall. Each one is a rule we can be held to, and each one costs us something.
Own the network, not just the app
We run our own SBCs, media servers and carrier interconnects. It is slower and more expensive than reselling, and it is the only way to fix a route rather than report it upstream and wait.
Publish the real numbers
Availability, MOS, ASR, post-dial delay and P1 response times are measured from production traffic and shown in your portal — including the months where a figure moved the wrong way.
No surprise billing
Published per-destination rates, no activation fees, no charge for porting standard ranges, and spend caps set at provisioning. If a rate changes, you get 30 days' notice in writing.
Engineers answer the phone
Our first line is staffed by voice engineers who can read a SIP trace, not by scripted agents who escalate. Median first response on a P1 is four minutes, at 03:00 as well as at 15:00.
Interoperate rather than replace
We maintain validated profiles for 60+ PBX and contact centre platforms. Keeping the Avaya estate you already paid for is a legitimate architecture, and we will support it instead of quoting a rip-and-replace.
Security is a default, not a tier
TLS 1.3 signalling, SRTP media, STIR/SHAKEN attestation A, fraud velocity guards and data residency pinning are included on every account. We do not sell encryption as an upgrade.
The people accountable for the platform
A small executive team, most of whom came from carrier operations rather than from software marketing.
Dinesh Kulkarni
Co-founder & Chief Executive
Spent nine years in Tier-1 carrier interconnect and settlement before Telecorenetwork, and still reviews the weekly route quality report line by line.
Marta Ribeiro
Co-founder & Chief Technology Officer
Wrote the first version of the media control plane. Leads switching, the CPaaS API and the interop lab, and owns the platform's latency budget.
Adaeze Okonkwo
VP Network Operations
Runs the 24/7/365 NOC across three follow-the-sun shifts and set the four-minute P1 first-response target the company now reports against publicly.
Henrik Thorsen
Chief Information Security Officer
Led the SOC 2 Type II programme and the ISO 27001 alignment work, and owns fraud detection, data residency and customer security reviews.
Team profiles shown here are illustrative placeholders for this template. Replace names, roles and biographies with your own before publishing.
Forty-two points of presence, one control plane
Every PoP runs the same software build, the same monitoring and the same routing logic. There is no premium region and no legacy region — a call anchored in São Paulo is handled by the same code path as one anchored in Frankfurt.
- Redundant SBC pairs in every PoP with automatic session re-anchoring in under 400 milliseconds
- Direct interconnects with 110+ national operators, plus private cross-connects for high-volume customers
- Anycast SIP edges so registrations land in the nearest healthy PoP without client-side configuration
- Route scoring on live MOS, ASR and ACD, with underperforming carriers demoted automatically
- Recording, CDR and metadata storage pinned to the EU, US, UK or APAC region you nominate
- Capacity reviewed monthly against measured busy-hour concurrency, not annual forecasts
Points of presence by region
Local, national, toll-free and mobile numbers are available across 8,000+ cities. Coverage lists are published in the customer portal and updated monthly.
Audited, aligned and documented
Audit reports, penetration test summaries and our data processing agreement are available under NDA — request them through the contact form or from your account engineer.
Working here
Around 240 people, roughly 60% of them in engineering and network operations. We hire for depth in a narrow field — SIP, media, routing, billing integrity, security — and we expect everyone to spend time close to production.
- Engineers take part in the on-call rotation and read their own incident reviews
- Remote-first across the US, UK, Germany, Singapore and Brazil, with four regional offices
- Structured 12-week onboarding including two weeks embedded in the NOC, whatever your role
- Published internal salary bands and no commission-driven support targets
Open roles are posted on our careers portal. If nothing fits but you have run a voice network at scale, write to us anyway at sales@telecorenetwork.com and we will route it internally.
~240 people
Across five countries, with hiring concentrated in network engineering and platform security.
60% engineering
Most of the company writes, operates or tests the software that carries your calls.
24/7/365 NOC
Three follow-the-sun shifts staffed by voice engineers, never outsourced to a call centre.
4 regional offices
London, Frankfurt, Singapore and São Paulo, alongside the United States headquarters.
Ask us the hard questions before you buy
Route quality on your specific destinations, how we handle a P1 at 03:00, what our last audit found, what happens to your numbers if you leave. Our engineers answer all of it directly, before any commercial conversation.
Written response within one business day · No sales sequence · Sandbox access on request