AlterTEL prices Polish local access through the routes it can actually repair

Summary

• AlterTEL is not a dormant name. Poland's current company and telecom registers show an active Warsaw company, its domain and telephone service remain live, and RIPE observations on 10 July 2026 showed seven IPv4 /24 prefixes originated by two AlterTEL autonomous systems.
• The two autonomous systems have materially different exposure. Every observed path to AS43674 entered through MetroLine's AS44655, while AS50366 was visible through both MetroLine and dcenter.pl's AS49895. That is routing diversity for one part of the estate, not proof of physically separate fibre, buildings, power or field crews.
• Public evidence does not support describing AlterTEL as the owner of a Warsaw fibre, fixed-wireless, pole or tower access network. UKE currently records retail fixed interpersonal communications using numbers, while AlterTEL's own site emphasises IP telephony, technical maintenance, electronics, telecom systems and data-centre work.
• AlterTEL's defensible advantage is consequently local operational knowledge rather than unqualified control of every route. A buyer should price separately the local loop, upstream exits, powered sites, customer equipment and restoration labour, and should require named owners, physical-route diversity and recovery commitments for each.

The sale begins where AlterTEL's spanner stops

A business circuit is often sold as a number: a monthly price, a committed bandwidth, a latency target or a block of telephone numbers. Failure is less abstract. A cable is cut in the street. A landlord will not open a riser room. A switch loses power. A voice gateway fails. An upstream withdraws a route. A replacement optical module is in another city. Each event has an owner, a location and a repair clock, and those three facts determine whether the service is back in minutes, hours or days.

AlterTEL's public proposition is built around intervention. Its website calls the company an alternative telecommunications provider and says its main activity is delivering technical services for demanding business customers. The listed offers are IP telephony, maintenance of technical installations, electronics service, audits, support for telecom systems and telephone exchanges, server installations, and the construction and maintenance of data-processing centres. The contact page gives a Warsaw commercial number and the same Nowogrodzka 51 address that appears in the state register.

This is a more revealing proposition than a generic promise of fast internet. It puts labour, fault diagnosis and customer equipment close to the centre of the offer. It also creates a strict limit. A technician can repair the equipment and cabling to which the company has access, but cannot instantly restore a leased metro fibre after a third party's civil-works accident, energise a carrier hotel during a wider power failure or force an upstream to reroute traffic. A local service provider is valuable precisely because it knows where that authority ends and has rehearsed the handoff to the next owner.

The distinction matters for the title's word access. There is historical evidence that AlterTEL delivered internet access in Warsaw: a 2018 UKE decision, using data for the end of 2017, listed AlterTEL among Warsaw internet providers and associated it with 1 Gigabit Ethernet. That entry establishes a service and technology at that time. It does not identify a street footprint, an owned fibre route, a current product, a wholesale supplier or a ring.

The current picture is narrower on the service side and stronger on the routing side. That combination argues for diligence rather than dismissal. AlterTEL can be a live operator without being the owner of every physical layer under its service. Its prices make sense only after the buyer identifies which failures AlterTEL can fix directly, which it can merely escalate, and how much supplier delay the contract assumes.

A live legal and regulatory footprint, but a narrow one

The legal entity has continuity. The Ministry of Justice's current KRS extract records AlterTEL Sp. z o.o. under KRS 0000337253, incorporated on 5 October 2009, with PLN150,000 of share capital and a Warsaw registered office. Its principal registered activity is wired telecommunications. The extract was current through 24 June 2026 and records annual financial filings through the year ended 31 December 2025. There is no liquidation, bankruptcy or deletion entry in the current sections.

Registration as a telecom business is also current. UKE's public register entry for NIP 7010194452 shows telecom-enterprise number 8614, entered on 9 October 2009 with no deletion date. The declared service is retail fixed interpersonal communication using numbers, including nomadic services, across Poland. The network-activity field says only "other". It does not currently list retail or wholesale fixed internet access, data transmission, leased lines, fibre, copper, coaxial cable or licensed or unlicensed wireless access.

That absence is not a finding that AlterTEL is prohibited from carrying IP traffic or maintaining customer networks. The register describes declared activities in regulatory categories, and its current row can differ from historical market data or a business-to-business arrangement delivered over another operator's access. It does set a boundary on what can be asserted from official service registration. The defensible current description is a numbered fixed-communications provider with active internet resources and technical-service capabilities, not a documented owner of a broad last-mile broadband system.

The history of telephone numbering supports real operation. UKE's 2012 decisions record a routing-number allocation for AlterTEL. Decisions from 2013 record Warsaw PSTN numbering and further routing numbers. In 2018, UKE recorded another Warsaw PSTN allocation. The following year's numbering decisions show AlterTEL returning or changing several numbering resources, including Warsaw and Kielce entries. Allocation and return are normal parts of operating a numbered service; neither discloses subscriber totals or the physical access used to reach them.

Number portability gives a more recent sign of customer movement. UKE's annual fixed-number portability report for 2025 records 15 numbers donated by AlterTEL, 502 received and a positive balance of 487. The prior 2024 annual report recorded a negative balance of 104. These are movements, not an installed-base count. A batch can reflect one business customer's number range, and porting says nothing about call quality, revenue or access ownership. The reversal nevertheless supports the conclusion that the numbered service was commercially active in 2025.

The public presence is maintained but sparse. AlterTEL's WordPress page record shows a homepage last modified in 2016 and a contact page updated in 2024. The domain itself remains registered: the Polish registry's RDAP record shows an original registration in August 2009 and a current expiry in August 2027. A long-lived, mostly static website is compatible with a small relationship-led business. It is not a substitute for a product catalogue, serviceable-address list, network map, status page or public service-level terms.

Two autonomous systems make AlterTEL visible

The clearest current operating evidence sits in the global routing table. RIPE's overview for AS43674 identifies the holder as AlterTEL and marks the system announced. Its routing status on 10 July 2026 showed two IPv4 prefixes, 512 addresses in total, visible to 326 of 327 RIPE RIS IPv4 peers. The announced-prefix record identifies 91.198.100.0/24 and 193.43.90.0/24. No IPv6 space was visible for the system.

AS50366 is larger. RIPE's overview also identifies AlterTEL and marks the ASN announced. Its current routing status showed five IPv4 prefixes, 1,280 addresses and visibility at 326 of 327 IPv4 peers. The prefix record lists five adjacent /24s from 109.95.136.0/24 through 109.95.140.0/24. It, too, showed no IPv6 announcement.

Together, the two systems originated seven /24s containing 1,792 IPv4 addresses at the observation time. Each of those seven origin-prefix pairs returned a valid Route Origin Authorisation in RIPE's validation service; examples for AS43674 and AS50366 show the cryptographic authorisation and a maximum length of /24. That is meaningful routing hygiene. It reduces the risk that networks applying origin validation will accept an unauthorised ASN for those covered prefixes.

The address resources are also tied to AlterTEL in the RIPE registry. The 109.95.136.0 to 109.95.140.255 record names ALTERTEL-NET, gives Polish country registration and identifies the AlterTEL organisation and maintainer. The 91.198.100.0/24 record is assigned provider-independent space under the same organisation. These records support resource control. They do not say whether the addresses terminate in one room, several data centres, customer buildings, hosted servers or network-address-translation pools.

An IP address count is not a capacity figure. A /24 carried over a congested 1Gbps uplink can perform worse than a small address pool on a well-engineered 10Gbps service. Conversely, a business-voice platform may need relatively little sustained bandwidth while requiring very low packet loss, jitter and interruption. The public routing table proves reachability and origin policy. It does not reveal port speeds, committed transit, oversubscription, traffic levels, router headroom, customer sessions or the number of active endpoints.

Nor do two ASNs necessarily mean two operationally independent networks. An organisation may retain separate ASNs because of acquisitions, address history, customer separation, different upstream arrangements or staged migration. AlterTEL does not publish why it uses both. For a buyer, the useful question is not why there are two numbers on paper, but whether a failure that removes one route, router, facility or supplier leaves the required service reachable through the other.

The routing exits are unequal

RIPE's observed-neighbour data draws the most important distinction. AS43674 had one observed neighbour: AS44655, operated by MetroLine. Looking-glass observations for 91.198.100.0/24 showed AS44655 immediately before AS43674 on every visible path at the research time. The same was true for 193.43.90.0/24. At the interdomain-routing layer, AS43674 was therefore single-homed through MetroLine.

AS50366 had two observed neighbours: MetroLine's AS44655 and dcenter.pl's AS49895. The looking-glass view for 109.95.136.0/24 showed both. Most collector paths entered through AS49895, while a minority entered through AS44655; the pattern was repeated across the five announced /24s. This is positive evidence of two routing-layer paths into AS50366.

The two providers themselves have broader connectivity. MetroLine's public peering record describes a regional network present at Equinix Warsaw, THINX Warsaw and TPIX. dcenter.pl's routing profile shows upstream connectivity to Orange Polska and RETN, while current observations also identify AlterTEL's AS50366 as a downstream. Those records explain how AlterTEL routes can reach the wider internet. They do not establish the contracted service, the port or location used by AlterTEL, or whether MetroLine and dcenter.pl reach AlterTEL over separate entrances.

Physical commonality is the unresolved risk. Two BGP sessions can ride fibres in the same duct, terminate in the same Warsaw facility, depend on the same meet-me room or share a local power feed. A backhoe cut before the paths diverge can remove both. A fire-alarm discharge, access-control failure or building evacuation can make both ports unavailable even when the upstream backbones remain healthy. A dual-homed routing diagram is therefore only the first half of a redundancy claim.

Several public technical clues suggest a close operational dependence on MetroLine, but they must be treated as clues rather than a disclosed contract. AlterTEL's domain uses ns1.mlnet.pl and ns2.mlnet.pl, its web address resolves inside the AS50366 block, and the certificate presented on the site's HTTPS endpoint at the observation time named billing.metroline.pl rather than AlterTEL. RIPE paths also place MetroLine next to both AlterTEL ASNs. This pattern is consistent with shared or supplied infrastructure. It cannot determine ownership, corporate control, physical co-location or the terms of any service agreement.

The evidence that would settle the matter is ordinary carrier documentation: the A-end and B-end of each upstream circuit, facility names, demarcation points, local-loop owners, route drawings, confirmation of diverse building entrances, and the identity of the power and remote-hands providers. For AS43674, a buyer should assume loss of AS44655 removes internet reachability unless a private or unobserved failover path is demonstrated. For AS50366, the buyer can credit observed upstream diversity but should not credit physical diversity until those documents show it.

The access plant is the missing asset

The assignment of IP resources and the presence of BGP routes prove control at layer three. They do not locate the cable between a customer and AlterTEL's first router. That first section is where most local failures happen and where the word local earns or loses its meaning.

AlterTEL publishes no current serviceable-address map, fibre-route length, pole count, tower list, wireless spectrum, access-node inventory or customer-building list. There is no public description of GPON, XGS-PON, point-to-point fibre, licensed microwave or unlicensed fixed wireless. The current UKE row does not declare fibre or wireless access technologies. The historical 1 Gigabit Ethernet listing says how a service was presented in 2017, not how the physical route was built or who owned it.

Several operating structures remain possible. AlterTEL could own short building or campus links and buy metro transport beyond them. It could provide voice and managed equipment over a customer's existing broadband. It could lease a complete circuit from another operator and add addressing, routing and support. It could use MetroLine or another provider for the local loop while retaining the customer contract. It could also combine these structures from site to site. Public evidence does not choose among them.

This uncertainty changes the recovery promise. On owned fibre, AlterTEL can hold splice closures, cable, optical modules and test equipment, dispatch a crew and control the repair sequence. On leased fibre, it can diagnose the likely fault, open a ticket, supply optical readings and press the carrier, but the field arrival and splice priority belong to someone else. On customer-supplied internet, AlterTEL may control only the voice gateway or tunnel. A low price can be rational in any of these structures, but each buys a different amount of control.

Poland's wider market makes such mixed structures normal. UKE's 2025 market report says more than 2,600 telecom enterprises operate in the country, most locally or regionally, and that micro, small and medium providers deliver more than half of fixed-internet services in rural areas. The report also expands its treatment of wholesale arrangements such as bitstream access and local-loop unbundling. A local service provider need not own national infrastructure to be useful. It must be clear about the portion it owns and the wholesale dependencies embedded in its offer.

An effective asset schedule would split every service into six fields. The first is customer-premises equipment: handsets, PBX, session border controller, optical terminal, router and UPS. The second is inside-building cable and access rights. The third is the local loop to the first aggregation point. The fourth is the powered router or voice platform. The fifth is upstream transit and interconnection. The sixth is the staff and spares needed to restore each layer. Without that schedule, "AlterTEL service" compresses several owners and several repair clocks into one label.

National authorisation is not national plant

The overview classifies this research in a global collection, but AlterTEL's evidenced operating geography is much tighter. The company is Polish, the registered office is in Warsaw, the UKE service area is the Republic of Poland, the historical internet-provider entry concerns Warsaw, and the announced address resources are registered in Poland. None of those facts supports physical infrastructure outside Poland. Even within Poland, only Warsaw has a specifically evidenced access technology.

The distinction between service reach and plant reach is especially important for IP telephony. AlterTEL's site says its IP-telephony service is independent of location. That can be true at the application layer: a configured handset or gateway may register over a suitable internet connection from another Polish city or from abroad. It does not mean AlterTEL owns the local access, can enter the building or can dispatch a nearby technician wherever the endpoint is used. Geographic portability transfers more responsibility to the customer's broadband, power and local support arrangements.

The telecom register's nationwide service area should be read in the same way. It says where the authorised service is offered, not where fibre has been installed. A provider can deliver a numbered service nationally by using wholesale interconnection and customers' existing connectivity. A national sales perimeter is therefore compatible with a small physical footprint. For recovery purposes, the relevant map is not the regulatory boundary but the location of the voice platform, internet exits, customer demarcations and people who can reach them.

IP registration is also not geolocation. RIPE records associate the resources with a Polish organisation, while route collectors see how the prefixes reach them from many vantage points. Neither identifies the room containing an AlterTEL router. Commercial IP-location services may place responsive addresses in Warsaw, but such locations can be inferred from registration, latency or network topology and are not proof of a street address. Facility identity should come from AlterTEL or the hosting and transit contracts.

For a Warsaw customer, the evidence is enough to make local attendance plausible but not guaranteed. For a customer elsewhere in Poland, the buyer should ask whether support is remote, subcontracted or dispatched from Warsaw, and whether the local loop is supplied by another carrier. For an endpoint outside Poland, the buyer should assume that AlterTEL controls the voice or routing service rather than the foreign access unless the contract names a local partner. This is how a geographically portable service can remain honest about the distance to repair.

Small-company economics favour service over speculative construction

AlterTEL's financial scale reinforces an asset-light interpretation, although it does not prove one. The official KRS extract confirms that 2025 accounts were filed. BizRaport's secondary extraction of those accounts reports PLN313,656 of 2025 revenue, PLN335,629 of total costs, a PLN21,973 net loss, PLN121,665 of assets and PLN208,195 of liabilities. It reports revenue of PLN318,615 in 2024 and PLN327,735 in 2023. These figures should be checked against the signed filing before credit decisions; the employment range on the same page is explicitly an estimate, not reported headcount.

Even with that caveat, annual revenue around PLN0.3 million is not consistent with assuming a large self-funded metropolitan fibre build inside this legal entity. A single civil-engineering project, carrier-grade router pair, annual transit commitment or round-the-clock staffed support desk can consume a meaningful share of that amount. The accounts may not capture assets owned by a supplier, customer or related company, and leased access can keep physical plant off AlterTEL's balance sheet. That is exactly why ownership must be demonstrated rather than inferred from the telecom label.

The company operates in a difficult product segment. UKE's 2024 market report found that fixed-telephony users fell 18.5 per cent in one year and revenue fell 13 per cent to PLN0.8 billion. VoIP was more resilient but still declined: 2.7 million users, down 2.8 per cent, and PLN293.7 million in revenue, down 0.4 per cent. Plain calling is not an easy growth engine.

The 2025 portability gain shows that a small provider can still win number blocks. The economic opportunity lies in the surrounding work: migrating a PBX, maintaining handsets and gateways, configuring failover, supporting a server installation, attending a site and giving a business one accountable contact. Those services reward local knowledge. They also expose the company to labour concentration. If only one person knows a customer's old exchange, VLAN plan or building route, sickness or simultaneous faults can become a capacity limit even when the network itself has spare bandwidth.

Pricing should therefore separate recurring transmission from human intervention. A cheap base service with chargeable site attendance may work for a customer that can tolerate delay. A hospital, contact centre, hotel or logistics site may need a higher fee that funds on-call coverage, stocked spares, a tested backup circuit and a firm restoration target. Bundling both into one undifferentiated monthly number makes the low-risk customer subsidise the difficult site or leaves the operator underfunded when several faults arrive together.

The more AlterTEL relies on upstream access, the more working capital also matters. Suppliers may bill fixed monthly port and loop charges while customer receipts arrive later. A major repair can require overtime or replacement equipment before an insurance or customer claim is resolved. A small balance sheet does not make the company incapable, but it makes contract design important: supplier back-to-back terms, deposits for custom construction, limits on service credits and a realistic inventory of critical spares protect both sides from promises that cannot be financed during an outage.

Installed capacity is not usable capacity

The public record contains several numbers that look like capacity and are not. Seven /24 prefixes describe address space, not bandwidth. A historical 1 Gigabit Ethernet entry describes an interface technology, not a committed end-to-end gigabit. Two upstream ASNs describe routing adjacency, not two physically independent cables. A national service area in the telecom register describes authorisation, not a Warsaw-to-Poland access footprint.

Usable capacity is the minimum available across the path when the network is busy or impaired. For an IP-telephony customer, that path includes the local access, any encrypted tunnel, AlterTEL's voice platform, interconnection to other carriers and the customer's own LAN and power. For an internet customer, it includes the customer port, aggregation, upstream transit and the return path selected by other networks. Congestion at any one point can damage performance even if every interface is labelled 1Gbps.

Voice is particularly unforgiving of short impairments. A file transfer can slow down and recover. A voice call exposes loss and jitter immediately, and a routing reconvergence can drop an established session even when reachability returns quickly. A business buying AlterTEL for IP telephony should ask for packet-loss and jitter thresholds, not only availability. It should also test emergency-call routing, number presentation, inbound failover and what happens to calls when the customer site loses power.

No public AlterTEL material gives committed information rates, oversubscription ratios, upstream port sizes, utilisation peaks or failover convergence. There is also no public IPv6 announcement from either ASN. IPv4-only visibility is not an outage, and customers can reach IPv6 services through translation or upstream arrangements, but the lack of native visible IPv6 is a limitation worth clarifying for a current business network.

Capacity claims should be tested under failure, not only in normal operation. If AS50366 loses its preferred dcenter.pl path, can the MetroLine path carry the full peak load without congestion? If services in AS43674 depend only on MetroLine, can selected customers move to AS50366, and are their addresses, firewalls and voice configurations prepared to do so? If both ASNs terminate on one router pair, does a control-plane or power failure remove them together? The routing table cannot answer these questions because it observes reachability, not internal design or load.

The most useful evidence would be a controlled failover report. It would record the failure introduced, prefixes withdrawn, convergence time, packet loss, call survival, remaining utilisation and restoration time. Repeating the test for a local-loop cut, upstream loss, router failure and power isolation would turn a general redundancy claim into an operational fact. Until then, the seven visible prefixes are installed routing assets whose usable capacity remains undisclosed.

Six failures reveal the real service

An access cut

A cut between the customer and first aggregation point is the purest test of local control. If AlterTEL owns the cable, the buyer should know the route, splice responsibility, locate process, spare-cable location and target crew arrival. If another carrier owns it, the contract should name that carrier and pass through its fault priority and restoration commitment. A second circuit is useful only if it avoids the same duct, bridge, building entrance and street works.

The affected customer depends on the use. A voice-only site may lose inbound and outbound calls while other internet service continues. A site using AlterTEL for both access and telephony may lose all external communication. A multi-tenant building can concentrate many customers behind one entrance cable. The impact mechanism is not the nominal bandwidth; it is the number of services sharing the damaged segment.

A facility or power outage

Routers, voice gateways, switches, servers and cooling require electricity. BEREC's current network-resilience overview notes that many regulators require backup power for core or access networks and that operators generally equip core networks more fully than the edge. That distinction matters to a small provider. A central platform may have generator-backed power while a building switch, rooftop radio or customer router lasts only as long as a small UPS.

AlterTEL publishes no facility names, power design or battery runtime. A customer should ask for runtime by device, not a general statement that backup exists. The same exercise must include cooling, remote access, generator fuel and the ability of staff to enter the site during a citywide incident. It should also include the customer premises: IP handsets and an optical terminal do not work during a local blackout merely because the carrier core remains online.

Loss of MetroLine

AS43674's current routing exposure is straightforward. The observed internet path enters through MetroLine. Loss of that adjacency would remove the two prefixes from normal global reachability unless AlterTEL has a failover that was not visible during the observation or can rapidly re-originate them elsewhere. Customers using services exclusively inside those prefixes should not assume that the two-ASN estate protects them.

AS50366 is better placed at the routing layer because both MetroLine and dcenter.pl were visible. Yet the MetroLine clue recurs in domain services and web hosting, and dcenter.pl may still meet AlterTEL in the same metropolitan location. The recovery test is to disable each upstream in turn and then isolate the shared facility or local loop. Passing the first test demonstrates BGP failover. Passing the second demonstrates physical resilience.

A router, platform or configuration failure

System failures have historically caused a large share of major European telecom incidents. ENISA's incident analysis also highlights power cuts and underground cable damage as recurring causes. A small operator's most dangerous device may not be a long fibre span but a single router, session border controller, virtualisation host or configuration repository on which many customers depend.

Two upstreams do not help if both terminate on one failed chassis. Two routers do not help if they share a software defect or a mistaken configuration pushed to both. Recovery evidence should identify equipment pairs, software separation where appropriate, configuration backups, spare hardware, rollback practice and the time needed to rebuild a voice or routing platform from clean state.

A field-repair shortage

AlterTEL presents technical service as a core competence, but it does not disclose current staff numbers, shift coverage, subcontractors, vehicles, test equipment or spare holdings. The website's statement that human capital is important is a claim of intent, not a capacity measure. The third-party employment estimate is too uncertain to use as a headcount.

Local support has an advantage only when it is available. The buyer should ask who answers at 03:00, who can enter each facility, who is authorised to work on the access plant, and what happens when two customers fail at once. Named alternates and a subcontractor roster matter more than a generic 24-hour contact number. For specialised voice systems, the company should also identify who understands legacy PBX cards, number routing and customer-specific configuration.

Congestion or a demand spike

Congestion can resemble a partial outage: calls remain technically connected but become unusable, remote desktops freeze and packet retransmissions consume more capacity. The risk rises during failover because surviving paths carry the displaced traffic. It also rises when a customer suffers a denial-of-service attack or a backup job runs at the wrong time.

No public data shows AlterTEL's normal or failover utilisation. A business SLA should therefore define performance thresholds and measurement points, not merely link availability. It should state whether denial-of-service mitigation is included, whether voice receives separate treatment on the local access, and what capacity remains when the largest upstream path is unavailable.

Recovery is a chain of named obligations

The European Commission's 2024 communications-resilience assessment recommends assessing international interconnections, the redundancy of core internet infrastructure, supplier and managed-service dependencies, and physical stress testing. The principle scales down neatly to AlterTEL. The relevant unit is not a grand national backbone but the complete chain from one Warsaw office or customer building to the wider network.

For the local loop, the contract should name the owner, technology, demarcation and physical route. It should distinguish an owned cable from wholesale access and identify whether a second service uses a different operator and street path. For a wireless link, it should add tower or rooftop rights, spectrum, line of sight, weather margin and backup power. No evidence reviewed for AlterTEL supports assuming a current fixed-wireless system, so that option belongs in diligence only where a site design actually uses it.

For the upstream layer, AS43674 and AS50366 should be treated separately. AS43674 needs proof of an alternate exit or an explicit acceptance that MetroLine is the single observed route. AS50366 can show two observed providers, but should document separate local circuits, separate entrances and enough standby capacity. The customer should also know whether its addresses and services can move between the two AlterTEL ASNs or are fixed to one.

For facilities and power, recovery terms need runtime figures, generator arrangements, fuel priority, cooling, remote management and access rights. A carrier hotel can have excellent utility resilience while the customer's access switch in an office basement has twenty minutes of battery. The end-to-end service inherits the shortest runtime. Testing should therefore isolate the edge as well as the core.

For spares, the useful list is short and specific: customer router, voice gateway, optical module, power supply, switch, storage device, fibre closure and enough compatible cable to make a temporary repair. The list should identify where each item is stored and who can install it. A spare in a vendor warehouse is supply-chain resilience, not immediate restoration.

For people, the contract should state response and restoration separately. Answering a telephone in ten minutes is not the same as arriving at a site, and arriving is not the same as restoring service. AlterTEL's local-service positioning can justify a premium when it commits to all three with enough coverage to survive concurrent incidents. Without that commitment, "local" is geography rather than resilience.

Finally, service credits should be aligned with business impact but not mistaken for recovery capacity. A credit after an outage does not reconnect a call centre. The stronger remedy is a tested secondary route, automatic voice diversion, mobile backup, spare customer equipment and a clear escalation path into each wholesale supplier. Price those elements visibly, and the buyer can choose the level of continuity it actually needs.

The unofficial signals point to dependency, not failure

The public clues outside formal service disclosures are coherent: mlnet.pl name servers, a MetroLine certificate on the AlterTEL web endpoint, MetroLine adjacent to both AlterTEL ASNs, and dcenter.pl adjacent to AS50366. They suggest that AlterTEL's operating surface is closely integrated with a small Warsaw network cluster. They do not prove that the companies share ownership, a building, staff or physical routes. They also do not show that any dependency is poorly managed.

The website's HTTPS mismatch is similarly limited evidence. It is a current configuration problem on a public endpoint and may reflect shared hosting or a default virtual host. It does not measure packet loss, voice availability or the security of the routed network. Its significance is narrower: a company selling technical care would benefit from publishing a correctly configured secure site, current service descriptions, a status contact and clear escalation information.

Financial aggregators and routing inventories require the same discipline. The financial figures are useful because they are extracted from filed accounts, but a signed statement remains the definitive document and an employment figure marked as estimated is not evidence of crew size. BGP collectors are strong evidence of routes visible from their peers at a particular time, but private interconnections, backup sessions and physical paths can remain unseen.

Three disclosures would settle most of the uncertainty. First, a current product and service-area statement would say whether AlterTEL sells internet access, managed voice, maintenance or a combination at each type of site. Second, an asset-and-supplier map would identify local-loop ownership, facilities, upstream circuits and physical diversity. Third, a recovery statement would give power runtime, spares, staffing, test results and restoration targets. None requires publication of sensitive router addresses or customer names.

AlterTEL's credible product is controlled repair

AlterTEL has enough evidence to reject the label of a paper operator. It has an active legal entity, current telecom registration, recent number-porting activity, a live domain, two announced ASNs, seven widely visible IPv4 prefixes and valid route-origin authorisations. AS50366 has two observed upstream paths. These are concrete operating facts.

It does not have enough public evidence to support a broad claim of owned regional access infrastructure. No current fibre or fixed-wireless footprint, pole or tower dependency, route map, customer count, facility list, capacity figure, power runtime, crew roster or repair performance is disclosed. AS43674 is visibly single-homed, and AS50366's two routing exits have unverified physical separation. The current regulator entry is centred on numbered fixed communications rather than internet access.

That leaves a narrower but potentially durable place in the Polish business market. AlterTEL can combine numbers, routing resources, equipment knowledge and local attendance, buying or partnering for the access it does not own. In a declining voice market, the value is not the dial tone alone. It is the ability to understand the whole customer path, isolate the failed layer and make the right party move quickly.

The price should follow that control. A service over one wholesale loop and one observed upstream should be sold as such. A resilient service should fund a separate route, sufficient standby transit, independent power, spare equipment and more than one qualified responder. A managed-voice offer should include customer-premises power and call-diversion tests. Each addition costs money because it removes a specific failure domain.

AlterTEL's network evidence is therefore medium rather than strong. Current routing and regulatory records establish operation, while the physical access and recovery surface remains largely private. The company can improve that grade without pretending to own a national network. It needs to show the routes it can repair, name the routes it cannot, and let the customer buy the difference.