Rootlenses
Rootlenses is a business AI suite that answers data questions in plain English, runs inbound and outbound calls with voice agents, and connects AI models to internal systems under role-based access and audit controls. Pricing is quote-only.
What is Rootlenses?
Rootlenses is a business AI suite from a Panama-based publisher, sold as three complementary products on one platform.
Rootlenses Insight is the analytics piece. You ask a question the way you would ask a colleague — which products sold best this month, which customers have outstanding payments, how many Java engineers are free — and it turns that into an optimised SQL query against your company databases, returning an answer, a table and a chart in seconds. It connects to MySQL, SQL Server, PostgreSQL, Oracle and modern data warehouses, reads their schemas by itself to work out the context, and needs no change to the source system. Reports download as CSV, XLS, JPG or PNG.
Rootlenses Voice handles the telephone. It builds AI voice agents that answer inbound calls and run outbound campaigns: qualifying leads, booking appointments, chasing payments, following up, and transferring to a human when the conversation calls for one. Agents work in parallel across thousands of simultaneous calls, detect answering machines, validate numbers before dialling, and produce transcripts with sentiment analysis and engagement scoring. Contacts arrive through a CSV template or through ETL scripts you write yourself, and call scripts can be versioned behind approval workflows.
Rootlenses MCP is the governance layer. It is an on-premise server implementing the Model Context Protocol, installed as a Docker container inside your own infrastructure — 4 vCPUs and 8 GB of RAM recommended, provisioned with Terraform. Credentials are configured locally, the data flow never leaves your network, and Rootlenses itself only manages licence activation.
Across all three, the platform leans hard on governance: role-based access control down to individual databases, tables or metrics, audit logging of every query and access, multi-factor authentication, and encryption in transit and at rest. Deployment can be cloud, hybrid or on-premise depending on regulatory constraints. The contract carries a 99.9% monthly uptime SLA, a 30-minute response target on critical support, and service credits when it is missed.
The service is strictly B2B. The sectors the site foregrounds are banking, healthcare, retail, insurance, real estate, HR and collections. There is no self-service path: every plan routes to the sales team, and no price appears anywhere.
What it does
- Ask a business question in plain language and get an answer, a table and a chart back in seconds
- Query several company databases at once without writing a line of SQL
- Run inbound and outbound call campaigns with AI voice agents working in parallel around the clock
- Have a voice agent qualify leads, book appointments, chase payments and hand over to a human when needed
- Generate and download reports and charts as CSV, XLS, JPG or PNG files
- Set roles and permissions down to a database, table or metric, and audit every query that runs
- Deploy an MCP server inside your own infrastructure to connect AI models to internal systems
When to use Rootlenses / When not to
A quick filter to help you decide if Rootlenses is the right fit.
When to use Rootlenses
- Business teams that currently wait on IT or an analyst every time they need a number out of a company database
- Operations running high call volumes — collections, appointment confirmation, lead follow-up — that cannot keep adding headcount
- Regulated organisations in banking, insurance or healthcare that need role-level access, query auditing and a choice of hosting model
- IT and data leaders who want AI models connected to internal systems without pushing data out to an external cloud
- Companies juggling several database engines — MySQL, PostgreSQL, SQL Server, Oracle — who want one place to question them all
When not to use Rootlenses
- Individuals and consumers: the terms restrict the service to businesses and professional teams only
- Buyers who want to sign up and pay online — every plan routes to a sales conversation and no price is published
- Teams shopping for a customer-facing text chatbot, which the FAQ states plainly that Insight is not
- Developers expecting a documented public API: none exists, and webhooks appear only as a plan quota
- Procurement that requires a named third-party certification such as SOC 2 or ISO 27001 on the vendor's security page
How to use Rootlenses
A typical end-to-end flow, from setup to results.
- Request a demo or contact sales — there is no self-service sign-up and no published price
- Work through the vendor's four-step method: needs assessment, expert consultation, solution scoping, then deployment and team training
- For Insight, connect a database — MySQL, SQL Server, PostgreSQL or Oracle — using credentials your IT team provides
- Let the agent read the tables and schemas on its own; no manual training step is required
- Ask questions in the chat as you would ask a colleague, then download the result as CSV, XLS, JPG or PNG
- For Voice, define the agent's objective, script and decision rules, choosing between an AI Call and a pre-recorded Audio Call
- Load contacts with the downloadable CSV template, or write Load, Transformation and Output ETL scripts sharing global variables
- Set the calling window, the delay between attempts, and a ring timeout between 10 and 30 seconds — 20 is the documented recommendation
- Test on a small file of 10 to 50 rows before the full upload, checking numbers are in E.164 format
- For MCP, deploy the container in your own environment with Terraform, configure tools and permissions, then monitor from the Admin UI
Pros & Cons
Pros
- Three complementary pieces — analytics, voice and AI governance — under one platform and one contract
- No SQL skills needed on the business side, and no migration: it connects to the databases you already run
- Governance built for regulated buyers: granular RBAC, query-level auditing, MFA, encryption in transit and at rest
- MCP runs inside your own network, so credentials stay local and the data flow never leaves your infrastructure
- A published, quantified SLA: 99.9% monthly uptime, 30-minute critical response, service credits when it is missed
- A plain line on model training: platform data is not used to train public AI models without explicit consent
- No lock-in — cancel at any time without penalty, with thirty days to export your data afterwards
Cons
- No public pricing at all: the seven plans all say 'Talk to sales' and the billing toggle itself reads 'Custom price'
- Neither a free plan nor a free trial is offered anywhere on the site
- No third-party certification is named — no SOC 2, no ISO 27001 — despite courting banking, insurance and healthcare
- No developer API documentation is published, and no data processing agreement is offered or refused anywhere
- GDPR compliance is never claimed outright; the stated framework is Panamanian law, with no Article 27 EU representative
- Liability is capped at the greater of twelve months of fees or 1,000 USD, and unused call minutes expire each cycle
- The company named in the contracts, Rootlenses, Inc., is not the one in the site's copyright line, Rootstack, SA
Pricing & Plans
There is no free plan and no free trial. Rootlenses publishes no prices whatsoever: all seven plans across Insight and Voice carry a 'Talk to sales' button, and both the monthly and the annual billing options are labelled 'Custom price', so a quote must be obtained from the sales team. What is documented is the metering. Voice consumes 24 credits per minute or part-minute of a connected call and 5 credits per failed call, and once a plan's included minutes are exhausted, billing switches automatically to pay-as-you-go at the plan's standard per-minute rate rather than blocking calls.
- 3 users
- one database connection
- email support answered within 24 hours
- standard security features and standard reports. Price on request
- 10 users
- up to 3 database connections
- priority email and chat support within 8 hours
- advanced security features. Price on request
- unlimited users
- up to 10 database connections
- premium 24/7 phone support with a dedicated account manager
- full security suite
- advanced analytics and custom AI model fine-tuning. Price on request
- 1 simultaneous agent
- 12
- 500 credits
- 2 agents
- unlimited platform users
- answering-machine detection
- self-service onboarding
- 10 GB of recording storage
- 3 simultaneous agents
- 37
- 500 credits
- 6 agents
- priority chat support
- assisted onboarding
- 20 GB of storage
- 90-day retention
- 5 simultaneous agents
- 87
- 500 credits
- 15 agents
- dedicated account manager
- assisted onboarding
- 40 GB of storage
- four one-on-one sessions with a Solutions Engineer
- 10 simultaneous agents
- 125
- 000 credits
- unlimited agents
- dedicated account manager
- assisted onboarding
- custom recording space
- five sessions with a Solutions Engineer
Data, GDPR & hosting
A consolidated view of how Rootlenses handles your data.
GDPR overview
Rootlenses never claims GDPR compliance in so many words, and that silence matters. What it does publish is GDPR-shaped: the privacy policy, effective 15 January 2026, states legal bases in Article 6 terms — contractual necessity, legitimate interests, consent — and grants access, rectification, erasure, restriction, objection and portability rights, exercised at jdflorez@rootstack.com. International transfers are said to rely on standard contractual clauses, GDPR adequacy mechanisms or other lawful safeguards, and the terms promise breach notification without unreasonable delay and, where feasible, within 72 hours. The stated framework, however, is Panamanian law and international best practice, not the GDPR itself. No Article 27 EU representative is designated, no data protection officer is named, and no data processing agreement is offered anywhere on the site.
Who owns the data?
The terms draw a clean line. Everything you feed in — call recordings, transcripts, queries, uploaded files — stays yours; you grant Rootlenses only the licence it needs to deliver the service. Everything the platform is made of — software, AI models, protocols, templates, scripts, visualisations — belongs to Rootlenses, and you receive a limited, non-exclusive, non-transferable licence for internal business use. The security page repeats the commitment: customer information remains the customer's property, is never sold, and is not shared with third parties without authorisation. Each organisation keeps control of its own data, users, permissions and access policies, and has thirty days after termination to export.
Reuse rights
Within that licence, the customer organisation may reuse everything it puts in and everything the platform produces for it — reports, charts, transcripts, files exported as CSV, XLS, JPG or PNG — for its own internal business purposes, without asking Rootlenses first. Nothing in the terms requires attribution or prior approval for that internal use. What the customer may not do is copy, modify, redistribute or reverse-engineer the platform itself. On the vendor's side, collected data serves to deliver and improve the service, keep it secure, meet legal obligations and feed aggregated product research; personal information is not sold. Information processed within the platform is not used to train public AI models without the customer's explicit consent.
Data retention & training
Hosting summary
Rootlenses does not declare where customer data lives. The privacy policy says only that personal data may be transferred outside your jurisdiction, giving 'U.S. servers' as its example, protected by standard contractual clauses, GDPR adequacy mechanisms or other lawful safeguards. No hosting region is named and no residency guarantee is given. Infrastructure sub-processors are described generically as cloud infrastructure vendors, alongside the named OpenAI and Twilio. The counterweight is deployment flexibility: the FAQ states Insight can run in cloud, hybrid or on-premise environments to satisfy data-residency policies, and Rootlenses MCP installs entirely inside the customer's own infrastructure, where credentials are configured locally and the data flow never leaves the customer's network, with Rootlenses managing licence activation only. The public website itself resolves to Amazon Web Services in the United States, which describes the marketing site rather than the platform's data stores.
Things to keep in mind
Risks and trade-offs to weigh before adopting Rootlenses.
- No published price means no budget check before you spend sales-cycle time; every evaluation starts with a conversation you cannot skip
- The privacy policy states the AI's knowledge stops at the end of 2021 and disclaims liability for hallucinations, so outputs need verifying before decisions rest on them
- Trusting a natural-language answer without ever inspecting the query behind it is the quiet risk: the tool makes it easy to stop looking at the data
- Automated calling shifts consent and call-recording compliance onto you — the terms make the customer responsible for local law
- Two legal identities appear on the site, Rootlenses, Inc. in the contracts and Rootstack, SA in the copyright, so check which one you are signing with
- Liability is capped at 1,000 USD or twelve months of fees, whichever is greater, which is low for a system touching core banking or patient records
- The GDPR and legal contact is one named person's address on a third-party domain rather than a functional mailbox, a single point of failure for data-subject requests
Setup & Integrations
Technical difficulty
Uneven, and it depends which piece you buy. Insight is the easy end: connect a database in minutes, let the agent read the schema, ask questions — no code on the business side, though IT still has to grant credentials. Voice is a configuration job: scripts, calling windows, retries, and either a CSV template or ETL scripts you write yourself. MCP is an infrastructure project — a Docker container in your own environment, 4 vCPUs and 8 GB of RAM recommended, provisioned with Terraform. Onboarding is self-service on the entry Voice plan and assisted above it.
Deployment
Integrations
Supported languages
Behind Rootlenses
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Rootlenses.
Frequently asked questions
Do I need to know SQL to use Rootlenses Insight?
Which databases can Insight connect to?
How much does Rootlenses cost?
Is there a free trial or a free plan?
Will my data be used to train AI models?
Where is my data hosted?
What happens if I use more call minutes than my plan includes?
Can I cancel at any time?
Is Rootlenses GDPR compliant?
What does Rootlenses MCP need to run?
Should you pick Rootlenses?
Rootlenses makes a coherent case for a specific buyer: a mid-sized to large organisation, often in a regulated sector, that wants natural-language access to its own databases and automated calling at scale without handing either to an external cloud. The governance work is real — role-level access down to individual tables, audit logging of every query, MFA, and an MCP server that runs inside your own network with credentials that never leave it. The published SLA is specific and quantified, and the position on model training is stated plainly rather than buried.
The reservations are just as concrete. Nothing is priced. Seven plans across the two products all end at 'Talk to sales', and the billing toggle itself reads 'Custom price', so there is no way to size a budget before entering a sales cycle. No free trial, no free plan. For a vendor courting banking, insurance and healthcare, the absence of any named certification — no SOC 2, no ISO 27001 — is a gap procurement will notice, as is the missing data processing agreement and the fact that GDPR compliance is never claimed outright; the stated framework is Panamanian law. The company named in the contracts, Rootlenses, Inc., is also not the one in the site's copyright line, Rootstack, SA, which is worth clarifying before signing.
The product is young: the domain's first archived snapshot dates from July 2025 and the current policies took effect in January 2026. That is not a fault in itself, but it argues for asking harder questions during the demo — about certifications, about where data actually sits, and about which entity you are contracting with. For teams that get satisfying answers, conversational analytics, voice automation and on-premise AI governance in one place is an unusual combination.
- Choosing a selection results in a full page refresh.
- Opens in a new window.