Summary
- The strongest current fact is a routed network, not a current retail offer. RIPE collectors saw AS140113 originate two IPv4 /24s and one IPv6 /32 on 10 July 2026, with almost complete collector visibility and valid route-origin authorisations.
- The company was a Category C ISP licensee for the Aurangabad service area in the Department of Telecommunications list dated 31 January 2025. It is absent from the department's active list dated 28 February 2026, while a same-date surrender list also does not name it. That combination does not prove cancellation, renewal, transfer or continued authority; it leaves current licence status unresolved.
- TRAI reported only five to eight Riyaz broadband connections per quarter from 2021 through March 2024, then stopped naming the provider in the published ISP annexures reviewed through March 2026. Those figures cannot be reconciled with third-party estimates of thousands of users or with an unofficial claim of fibre plans up to 300 Mbps.
- Public routing views show only one observed adjacent network for both IPv4 and IPv6. Riyaz also declares operational route-server sessions at DE-CIX Mumbai and NIXI Mumbai, but exchange access is not a substitute for full internet transit and does not prove two physically independent paths from Aurangabad to Mumbai.
- There is no public route map, access technology inventory, fibre ownership statement, pole or duct agreement, power-runtime target, field-force disclosure, fault-repair metric, tariff, customer contract or current serviceability checker attributable to Riyaz. The safer conclusion is a live network-resource and routing story, not a verified current retail-ISP story.
A live autonomous system is not the same thing as a live retail ISP
Riyaz Internet presents an unusually useful warning against collapsing several different kinds of telecom evidence into one conclusion. A company can exist in corporate records. It can have held an internet-service authorisation. It can control number resources and originate routes. It can appear at internet exchanges. It can even respond to traffic measurements. None of those facts, alone or together, proves that a household can order a new line today, that an existing customer still receives a bill from the same legal operator, or that a crew will repair a severed access cable.
The routed part of the picture is current. The RIPEstat routing-status view for AS140113 recorded a last-seen route on 10 July 2026. It saw 326 of 327 IPv4 collector peers and 320 of 321 IPv6 collector peers receiving Riyaz-originated space. Its announced-prefixes view listed 103.151.16.0/24, 103.151.17.0/24 and 2400:3ce0::/32 throughout the preceding two-week observation window. This is strong evidence that AS140113 is globally reachable and actively originated. It is not merely a dormant registration.
The retail part is much less secure. The Department of Telecommunications' active ISP list as at 31 January 2025 included Riyaz Internet Services Pvt Ltd under licence DS-11/139/2018-DS-III, Category C, for Aurangabad. The department's active list as at 28 February 2026 does not contain the company or that licence number. The department's same-date surrendered-licence list does not name Riyaz either.
Absence from one active list is a material negative signal, particularly because the older row was clear and specific. Yet absence is not a licence decision. It does not say whether the authorisation expired, was terminated, was transferred, was under amendment, was omitted in error or was replaced by some other permission. The surrender list closes only one possible explanation. A current licence letter, a direct entry in another official authorisation list or a reasoned department order would be needed to settle the matter.
The distinction matters to the title. The physical proposition is sound: any local fixed connection depends on access plant, power, upstream reachability and repair. But the public evidence does not establish that Riyaz currently sends such a bill. This article therefore evaluates the infrastructure that can be seen, the local service that was previously reported and the information required to connect the two. It does not present an unverified current consumer offer as fact.
The identity converges on Aurangabad, with one important naming difference
The assigned name is Riyaz Internet Service Private Limited, using singular "Service". Official telecom and corporate references generally use "Riyaz Internet Services Private Limited", in the plural. The convergence of identifiers makes it much more likely that these are naming variants for the same company than separate organisations.
The 2025 DoT row names Riyazuddin Faiyazuddin Shaikh as director, gives a Sardar Tower address in Aurangabad, and dates the Category C authorisation to 19 July 2018. A TRAI-hosted copy of the February 2024 authorisation list gives the same licence number, director, service area, address and effective date. An older Bharat Broadband Network copy of the ISP authorisation list describes the service area more fully as Aurangabad SSA in Maharashtra. These are independent official publication surfaces for substantially the same licence record, not three independent operating tests.
Corporate aggregators drawing on Indian company filings identify CIN U74999MH2017PTC303151, incorporation on 20 December 2017 and a registered address at 5-2-39, Sardar Tower, Osmanpura, Aurangabad. ZaubaCorp's company page names Riyazuddin Faiyazuddin Shaikh and Naushad Isak Baig as directors and reports the company as active in its captured filing view. The Company Check reports a March 2024 balance-sheet filing and a September 2024 annual meeting. These are useful identity and continuity signals, but they are secondary presentations rather than a current certified company extract. More importantly, an active private company is not necessarily an authorised public ISP.
The network records converge on the same place. PeeringDB's organisation page gives an Aurangabad address in New Usmanpura and links the organisation to AS140113. The APNIC RDAP record for AS140113 names RISNET-AS-IN, places it in India, records registration on 4 May 2020 and reports the number resource as active. The APNIC record for the IPv4 allocation names RISNET and covers 103.151.16.0 through 103.151.17.255. The IPv6 record likewise names RISNET.
There is even evidence of a commercial dispute surface. A Telecom Disputes Settlement and Appellate Tribunal cause list for 19 August 2025 lists a petition by Tata Communications Limited against Riyaz Internet Services Pvt Ltd and others. A cause list proves only that a named matter was scheduled; it does not reveal the claim, the contract, the outcome or whether the dispute concerned transit. It is nevertheless another dated appearance of the legal name in the telecom sector.
The responsible identity conclusion is narrow. The corporate, licence, address and network records refer to one Aurangabad company using singular and plural variants of "Service". The legal display name should follow the plural form used in the licence and corporate records, while the existing assigned entity remains the only linked company here. Nothing in this evidence supports inventing a parent, subsidiary, successor or ownership relationship.
The subscriber trail is tiny, stable and then silent
TRAI's quarterly annexures provide the only repeated public measure found for Riyaz's customer scale. They are also the clearest reason not to treat address-space size or third-party user estimates as retail subscriber counts.
The July-September 2021 report listed five Riyaz broadband subscribers. The 2021 yearly report listed eight. The January-March 2022 report listed seven, as did the April-June 2022 report, the October-December 2022 report, and successive reports through 2023. The October-December 2023 report still showed seven. The January-March 2024 report again showed seven.
Those are single-digit connections, not thousands. TRAI's annexures label the columns as subscriber numbers and elsewhere present large operators with full integer totals. There is no stated thousands multiplier for these rows. The reports also warn that their statistics are mainly reliant on information furnished by service providers. The values should therefore be read as regulator-published, provider-reported connections, not a field census.
Seven connections could describe a very small retail base, a small number of institutional or wholesale accounts, incomplete reporting, a dormant commercial operation, or an accounting boundary in which local users sit behind another reported connection. The annexure does not say which. It would be a mistake to multiply seven by an assumed household size or to infer an access topology from it.
The sequence then breaks. Riyaz is not named in the April-June 2024 report, the July-September 2024 report, the October-December 2024 report, or the reports reviewed for the four quarters of 2025. It is also absent from the ISP-wise subscriber annexure in the January-March 2026 report. The reports do not provide a note explaining why a previously listed provider disappeared.
That silence must be handled carefully. It does not prove zero subscribers, because reporting populations, presentation thresholds and provider identities can change. It also cannot be treated as evidence of growth. The clean statement is that TRAI last published a named Riyaz count of seven for March 2024, and the later reviewed annexures do not name it.
Other public estimates are not substitutes. Cloudflare Radar's AS140113 page has shown an estimated population of about 8,000 users, while an APNIC Labs measurement page estimated a population above 10,000 in a December 2024 observation. Such measurements infer users from internet observations; they do not identify who pays Riyaz, whether addresses are shared, whether customers are wholesale or retail, or which legal operator holds the contract. Carrier-grade address sharing, institutional networks, downstream arrangements and sampling assumptions can all separate estimated users from billed subscriptions.
An unofficial Aurangabad broadband blog offers the opposite kind of signal. Its 2023 provider roundup associates Riyaz with fibre-to-the-home and plans up to 300 Mbps. The page provides no dated tariff document, service terms, coverage map or evidence that it was written by the company. It suggests that Riyaz was known locally as a consumer-facing fibre brand. It cannot prove that the offer was accurate, generally available or still sold.
The conflict is the finding. Single-digit regulator-reported connections, thousands of statistically estimated users and an unofficial 300 Mbps retail claim describe different things. A current subscriber return, customer invoice, tariff sheet and address-level serviceability result would be required to reconcile them. Until then, subscriber scale and active retail status remain weak evidence.
AS140113 is small, dual-stack and cleanly authorised
The internet edge is easier to describe than the access business. APNIC assigned AS140113 in May 2020. It also allocated one contiguous IPv4 /23, equivalent to 512 addresses, and an IPv6 /32. The current global routing table carries the IPv4 space as two /24 routes and the IPv6 allocation as one /32 route.
This difference between allocation and announcement is normal but important. An allocation defines the address space held by the registered network. A BGP announcement tells other networks how to reach a prefix at a particular time. Dividing a /23 into two /24 announcements can offer routing-policy flexibility, but it does not prove two physical access networks, two routers, two upstream circuits or two cities. Both /24s can traverse the same port, fibre and provider.
Route-origin security is a positive. The RIPEstat validation result for 103.151.16.0/24 and the corresponding result for 103.151.17.0/24 show both routes covered by a valid authorisation for AS140113, with a maximum length of /24. The IPv6 validation result is also valid. This lowers the chance that networks enforcing route-origin validation will reject the announcements as unauthorised. It does not prevent cable cuts, configuration errors, congestion or a validly originated route leading to an unavailable service.
The route history also resists an easy story. RIPEstat reports its first observed Riyaz-originated prefix on 30 April 2023, almost three years after the ASN registration and after the earliest TRAI subscriber entries. That does not mean Riyaz lacked connectivity before 2023. It may have used provider-assigned space, routes not visible to the relevant collectors, or a different operating arrangement. The public evidence simply does not connect the pre-2023 subscriber reports to a documented AS140113 path.
Hurricane Electric's current AS140113 view shows the same two IPv4 routes, one IPv6 route, valid origin status and a single observed adjacent ASN on both protocols. The agreement across APNIC, RIPE collectors and this independent route view makes the present origin evidence strong. It still says little about customer premises. An autonomous system is an administrative routing domain; it can support retail broadband, enterprise connectivity, wholesale service, hosting or a mixture.
IP-level observations add modest evidence of use. IPinfo's AS140113 page has displayed responsive addresses and a recent traceroute reaching the network from a Mumbai probe. It classifies the network as an ISP and reports consumer-like daily activity in some snapshots. Such classification is inferred. A responsive address proves that an endpoint answered a measurement, not that a household line was provisioned, that the whole /23 is in use or that service quality is acceptable.
The safe conclusion is that Riyaz controls a modest but current dual-stack routing footprint with good route-origin hygiene. This is meaningful operating evidence. It is limited public evidence to upgrade the company to a verified current retail broadband operator without the missing regulatory and customer-facing evidence.
One observed neighbour and two exchanges are three different facts
The most consequential routing fact is not the address count. It is the narrowness of the publicly observed path out.
RIPEstat's ASN-neighbours result shows one adjacent network, AS137633, on the left side of observed paths. Hurricane Electric also reports AS137633 as the sole IPv4 and IPv6 peer visible for AS140113. In ordinary path interpretation, that position is consistent with AS137633 carrying Riyaz toward the wider internet. It is still an inferred relationship from observed routes, not a published contract. Route collectors can miss private sessions, backup paths that are not selected, and exchange routes that do not propagate in the same way.
A separate commercial dataset has at times classified a different ASN as Riyaz's only upstream. That disagreement reinforces the need for caution. It may reflect observation time, a provider change, incomplete collectors or classification rules. The current agreement between RIPEstat and Hurricane Electric supports one presently visible neighbour, but not a claim that Riyaz has never had another provider or lacks a cold standby.
PeeringDB records two exchange connections. Its public network API entry for AS140113 declares an operational 3 Gbps port at DE-CIX Mumbai with IPv4 and IPv6 addresses and an operational 1,024 Mbps port at NIXI Mumbai with IPv4. Both are marked as route-server sessions. The organisation last updated the network entry in May 2023, so the "operational" flag is entity-maintained rather than a live port test.
The two exchange locations are valuable if the sessions and transport remain live. They can let Riyaz reach content networks, cloud providers and other access networks without sending every packet through paid full transit. They can reduce latency and transit cost for routes learned at the exchanges. They can also provide two exchange ecosystems rather than one.
But exchange participation does not automatically provide the default-free reach of the entire internet. A route server distributes participating members' routes according to policy; it is not necessarily a transit provider for destinations beyond those members and their advertised customers. If the one observed full-reach neighbour fails, exchange-learned routes might keep some destinations available while everything else disappears. The exact surviving set depends on sessions, filters, local preference, default routes and transport that are not public.
Nor do two exchange names prove physical diversity. Riyaz is based in Aurangabad while both declared exchange sessions are in the Mumbai metro. The public record does not identify how it reaches those ports. Both could be delivered over one carrier circuit, one leased wavelength, one shared aggregation network or two genuinely independent paths. They could terminate in one facility or separate facilities. PeeringDB lists no Riyaz facility entries, and no carrier, circuit identifier or meet-me-room information is disclosed.
This is the difference between logical and physical redundancy. Logical diversity means separate sessions, exchange fabrics, IP addresses or routing policies. Physical diversity means that a single road excavation, duct fire, pole failure, carrier outage, power event or building fault cannot remove all of them. The first is partly visible. The second is not.
The declared port rates must also stay in their lane. A 3 Gbps exchange port plus a 1,024 Mbps exchange port does not mean customers have 4.024 Gbps of usable internet capacity. Exchange ports carry only exchange-learned traffic, may be rate-limited elsewhere, and may share the transport from Aurangabad. PeeringDB also self-reports a 5-10 Gbps traffic band and unusually high prefix-count values that do not match the three current originated routes. Those stale or differently defined fields show why entity declarations should be checked against live routing and measured utilisation.
For resilience, the decisive unanswered questions are concrete. Is there a second full-transit contract? Does it enter Aurangabad on a separate carrier and route? Are the NIXI and DE-CIX handoffs in separate facilities? Can either exchange transport survive the loss of the other? How much paid-transit capacity remains if all exchange routes are withdrawn? No public source found answers them.
The last mile is the largest blank space
The title's reference to field repair begins where BGP evidence ends. A packet from a customer must first reach Riyaz's routing domain. That requires some combination of customer equipment, drop cable or radio, local distribution, aggregation, a powered point of presence and backhaul. The sources establish none of those components as current Riyaz-owned assets.
The unofficial 2023 blog's fibre-to-the-home description is a useful lead, not an engineering record. If accurate, a typical path might include an optical network terminal at the premises, a fibre drop, a passive splitter, feeder fibre and an optical line terminal at a powered site. But no Riyaz source confirms a passive optical network, equipment vendor, split ratio, feeder design or terminal ownership. The image of that chain should therefore remain explanatory, not a claim about a specific deployment.
Fixed wireless is also possible in the abstract, especially for a small regional operator, but no source found identifies Riyaz towers, licensed spectrum, unlicensed radios, microwave backhaul or rooftop sites. The assignment's broad asset hypothesis cannot be narrowed responsibly to fibre, fixed wireless or a hybrid. Even the phrase "last mile" must not be confused with ownership: a provider can sell service over leased fibre, a local cable operator's access plant or another operator's wholesale connection.
The boundary matters when service fails. If Riyaz owns the drop and feeder, it controls dispatch and splicing but must fund vehicles, test sets, closures, spare cable and trained labour. If a local partner owns the access plant, Riyaz may control the bill and routing while the partner controls restoration. If a wholesale operator owns the access circuit, Riyaz's support team may only open a ticket and escalate. Without terms of service or an operator map, the accountable party at each layer is unknown.
Power has similarly different boundaries. A passive fibre splitter may need no local electricity, but the customer's terminal and Wi-Fi router do. The optical line terminal, aggregation switch and edge router do too. A fixed-wireless path adds powered radios at intermediate sites. Backup at a central site cannot keep a customer's unprotected terminal alive, and a household inverter cannot restore a failed upstream cabinet. No source discloses Riyaz battery runtime, generator coverage, fuel arrangements, maintenance tests or power design.
Aurangabad's urban setting creates both opportunity and exposure. Dense demand can shorten drops and spread fixed costs across more customers. Roads, building permissions and existing utility corridors can also create shared-fate routes. Fibre on two cables is not diverse if both occupy the same duct or pole line. A ring does not protect service if both sides converge at one bridge, cabinet, building entry or power supply. Public route counts cannot reveal any of this.
The missing field evidence is especially important because TRAI's current performance regime includes fault incidence, service provisioning, customer-care accessibility, latency, packet loss, jitter and maximum bandwidth utilisation. The January-March 2026 performance report explains benchmarks including no more than five wireline faults per 100 subscribers, broadband latency no more than 50 ms, packet loss no more than 1%, and customer-serving-node link utilisation no more than 80%. Riyaz does not appear in the provider tables reviewed, so those industry benchmarks cannot be presented as its results.
No current Riyaz fault rate, mean repair time, next-working-day result or major-outage history was found. That makes the planned Local support labour topic too strong. Field labour is physically necessary for a fixed access network, but its existence, employment model, coverage and performance are not established for this company.
A failure can be local even when the route remains visible
The first failure path is at the premises. A cut drop, damaged connector, failed optical terminal, misaligned radio or unpowered router can disconnect one user while AS140113 remains perfectly visible worldwide. A remote routing monitor would report a healthy origin. The customer would report no service. Restoration would depend on diagnosis, access to the building, a replacement unit or a technician capable of testing optical power or radio alignment.
The second path is the distribution segment. A feeder cut can remove a street or neighbourhood. If the access system is passive optical, one damaged feeder can affect all users behind a splitter tree; if it is active Ethernet or wireless, a failed powered node can have a similar footprint. The number affected depends on design and take-up, neither of which is public. A spare route helps only if the topology supports switching around the break.
The third path is power. A short customer outage may be solved by a small uninterruptible supply at the premises, but a longer area outage tests batteries and generators at each powered network site. Battery runtime is installed capacity; actual recovery depends on battery age, load, ambient conditions, alarms, access and fuel. Nothing in an ASN record or exchange profile measures it.
The fourth path is aggregation congestion. Even with intact fibre, too many busy users or limited public evidence uplink capacity can reduce throughput and raise latency. The two /24 routes do not reveal subscriber contention. The 5-10 Gbps PeeringDB traffic band, last updated in 2023, does not reveal peak demand in 2026. The exchange port rates do not reveal transit capacity. A 300 Mbps plan mentioned by an unofficial blog is an access ceiling, not a promise that 300 Mbps survives simultaneous demand or an upstream failure.
The fifth path is the Aurangabad-to-Mumbai transport. The public exchange records imply that traffic can meet other networks in Mumbai, but not how it gets there. A single leased circuit could be a common point of failure for both exchanges and the observed upstream. A physically diverse pair could protect against one cut while remaining vulnerable to a common building or power system. Only route diagrams, carrier letters and failover tests can distinguish these designs.
The sixth path is upstream loss. The current route collectors show one neighbour. If that neighbour withdraws Riyaz's prefixes and no alternate full route is available, outside networks lose reachability. Exchange sessions may preserve direct routes to participating networks, but not necessarily the rest of the internet. If the same transport carries both transit and exchange traffic, a single circuit failure can remove all three logical relationships at once.
The seventh path is repair saturation. One isolated cut can be easy to fix; a storm, road project, utility failure or multiple simultaneous breaks can exhaust technicians, vehicles, splicing equipment and spare cable. Average repair time hides this tail. Useful evidence would include median and 95th-percentile restoration, crew coverage by shift, spare-stock targets and performance during multi-fault incidents. None is available for Riyaz.
The eighth path is organisational uncertainty. If the retail authorisation or customer contracts changed while AS140113 remained active, the company responsible for answering a support call may not be obvious from routing records. A customer needs a named contracting entity, grievance contact and service terms. A network buyer needs the operator responsible for the ASN, circuits and exchange sessions. Those may be the same party, but the public evidence no longer proves the whole chain.
Different users feel these failures differently. A household loses work, study and entertainment. A shop may lose cloud point-of-sale access. A small office can lose voice and remote systems. A downstream local operator, if one exists, would magnify the affected population behind a single Riyaz connection. The reference does not identify customer classes or downstream networks, so these are exposure types, not attributed customers.
The sequence explains why route availability is necessary but not sufficient. A healthy BGP origin cannot repair a drop. A fast crew cannot reach the internet if the only upstream path is gone. Resilience requires both, joined by power and transport that do not share one hidden failure point.
Installed capacity is not surviving capacity
Public network profiles tempt readers to add numbers. Riyaz has 512 IPv4 addresses, a very large IPv6 allocation, a 3 Gbps declared DE-CIX port, a 1,024 Mbps NIXI port and a self-declared 5-10 Gbps traffic band. None can be summed into a customer-speed or resilience claim.
Address space is not bandwidth. IPv6 scale is designed to make address conservation unnecessary; a /32 says nothing about attached users or traffic. IPv4 can be used sparsely, shared, reserved or assigned to infrastructure. The two routed /24s show reachability, not utilisation.
Port speed is a physical or configured ceiling at one handoff. Actual exchange traffic depends on routes, peers, transport and demand. A port may be underused. It may also be fed by a narrower backhaul. The entity-maintained record may lag a change. NIXI and DE-CIX traffic cannot automatically substitute for transit because they expose different destinations.
Plan speed, if the unofficial 300 Mbps claim was ever accurate, is another ceiling. Customer throughput is constrained by the access link, aggregation, Wi-Fi, destination, peering and transit. A retail plan can remain technically provisioned while upstream congestion makes it unusable. Conversely, a small address block can support many customers through address sharing.
Surviving capacity is the amount left after the largest credible failure. If one 10 Gbps transport carries every logical connection, surviving capacity after that transport fails is zero. If two independent routes each carry 2 Gbps, surviving capacity may be 2 Gbps, subject to power and routing. The public sources do not disclose Riyaz's normal peak, headroom, failover policy or post-failure capacity.
This is also why the economics cannot be reconstructed. There is no current tariff, installation fee, equipment deposit, churn figure, revenue disclosure, customer count or network-cost breakdown. Regional ISP economics usually turn on density, take-up, support labour, access-plant cost, transit, peering and the capital required for redundancy. For Riyaz, only parts of the interconnection surface and an old single-digit subscriber series are visible. The planned Regional ISP economics topic would invite conclusions the sources cannot support.
What would upgrade the evidence
The highest-value document is a current authorisation record. It should identify the legal licensee, category, service area, effective dates and any amendment, transfer, surrender or termination. Because the February 2026 active and surrender lists both fail to explain Riyaz's status, a direct department record matters more than another company-profile page.
The second is a current customer surface: a dated tariff, terms of service, serviceability result, recent invoice or public support page naming the contracting entity. This would establish whether Riyaz still sells local broadband and whether the legal name on the bill matches the operator of AS140113.
The third is a subscriber and service mix reconciliation. A current TRAI return or company disclosure should distinguish retail lines, enterprise circuits, wholesale links and users behind shared connections. It should explain why published counts remained between five and eight through March 2024 and then disappeared while measured-user estimates remained much higher.
The fourth is an access and responsibility map. It need not expose sensitive street-level routes. A useful disclosure would state whether the offer is fibre, fixed wireless or mixed; which equipment the company owns; whether local partners install and repair; and where responsibility passes to a carrier, building owner or customer.
The fifth is physical diversity evidence. Riyaz could identify the number of full-transit providers, the number of independently routed Aurangabad exits, whether exchange and transit paths use separate carriers, and whether Mumbai handoffs occupy separate facilities. A redacted third-party diversity attestation would be more persuasive than adding another BGP session over the same path.
The sixth is operational performance. Median and tail repair times, significant outage minutes, packet loss, latency, utilisation, customer-care accessibility and power-backup targets would make field resilience measurable. Dates and definitions matter; a quarterly average should not be presented as proof of performance during a major failure.
Until those records appear, buyers should ask narrow questions. Which legal company signs the contract? Under which current authorisation? What medium reaches the premises? Who owns the drop? What powers the customer terminal and nearest active node? How many physically independent routes leave Aurangabad? Which full-transit provider takes over if the observed path fails? Which destinations remain reachable through the exchanges? Who dispatches a technician, and on what timetable?
These questions do not presume wrongdoing or failure. They are the minimum needed to turn live routes into an accountable access service.
Evidence-bound conclusion
Riyaz Internet has a real and current network footprint. AS140113 originates two IPv4 /24s and one IPv6 /32 with valid route-origin authorisations and near-complete visibility across RIPE collectors. Its entity-maintained PeeringDB profile declares route-server access at two Mumbai exchanges. The network is too active to dismiss as a dead registration.
The retail and resilience claims stop well short of that confidence. The last clear active licence row is dated January 2025. Riyaz is missing from the February 2026 active list and is not explained by the same-date surrender list. TRAI last published a named subscriber count of seven in March 2024 and did not name the provider in the later annexures reviewed. The company's own current tariff, coverage, support and service terms were not available. No public evidence verifies access technology, route ownership, independent backhaul, second transit, power runtime or field-repair performance.
The final network evidence grade is Weak. That grade does not deny the active ASN. It marks the gap between a live routing domain and the assigned story of a current regional retail ISP. The title's infrastructure logic remains conditional: if Riyaz is billing local fixed-connectivity customers, those bills depend on an access path, powered equipment, a narrow publicly observed route out and field repair. Public sources do not currently prove that commercial premise.
The metadata should follow the evidence. India, specifically Aurangabad in Maharashtra, is the supported geography, not Global. The official legal name uses "Services" in the plural. The regional-ISP category and retail-bill title should be held or reframed until current authority and customer operation are confirmed. Of the planned controlled topics, only Peering and transit is sufficiently supported by current company-specific evidence.