ThoughtSpot Modeling & Connectivity
Modeling & Connectivity is ThoughtSpot’s governed semantic foundation — define the model as code with TML (Git-friendly, version-controlled) and connect live to your cloud data warehouse with no data movement. The layer that makes Spotter’s AI answers trustworthy.
Buy through TechBag
Same software. Better outcome — at a lower cost.
How it’s rated
Full scoreboard ↓Quick answer
The ThoughtSpot platform family
This page covers Modeling & Connectivity — the semantic foundation. The rest of the ThoughtSpot platform:
Quick facts
30-second orientation- Product
- Modeling & Connectivity — TML + live query
- Vendor
- ThoughtSpot (founded 2012 · CEO Ketan Karkhanis)
- The layer
- The governed semantic foundation for AI answers
- What it does
- Code-first semantic model (TML) + live warehouse query
- TML
- Git-friendly, version-controlled data models as code
- AI-assisted
- SpotterModel builds & enriches the model
- Connects to
- Snowflake, Databricks, BigQuery, Redshift, Synapse
- India
- Indian-origin founders · huge Bangalore R&D
- Vs
- Looker (LookML), dbt Semantic Layer, Cube, AtScale
- In India via
- TechBag — scoping, model help, local support, GST
Understand code-first semantic modeling before you buy it
Most product pages skip this. We start here — so you buy a capability, not a buzzword.
What is Modeling & Connectivity?
ThoughtSpot’s governed semantic foundation — define the model as code with TML (Git-friendly, version-controlled) and connect live to your cloud warehouse with no data movement. The layer that makes Spotter’s AI trustworthy.
Click-built modeling vs code-first governed modeling (TML) — the honest table
What consolidation actually replaces, dimension by dimension.
| Dimension | Click-built modeling | Modeling & Connectivity |
|---|---|---|
| The semantic model | Click-built, no history | Code-first (TML), Git-friendly |
| Change control | Ad-hoc, undocumented | Review, diff, promote (DevOps) |
| Metric definitions | Different per team/dashboard | Certified once, one source of truth |
| Data | Extracts / stale copies | Live on the cloud warehouse |
| Warehouses | One, or copy in | Snowflake/Databricks/BigQuery+ live |
| Modelling effort | All manual | SpotterModel — AI-assisted |
| AI answers | Ungrounded, hallucinate | Grounded in the governed model |
| Best fit | (varies) | Code-first governed modeling on the modern stack |
ThoughtSpot Modeling & Connectivity is the governed semantic foundation — code-first, version-controlled modeling with TML (Git-friendly), live query on your cloud warehouse (Snowflake/Databricks/BigQuery+) with no data movement, and SpotterModel for AI-assisted modeling. It’s what makes Spotter’s AI answers trustworthy. Honest: premium/quote-priced, cloud-DW-oriented, TML has a learning curve. Benchmark modeling? Looker/LookML. Headless layer? dbt/Cube. OLAP? AtScale. TechBag scopes it, helps the model & adds GST.
The five pieces of the platform
Vendors love diagrams; buyers need to know what they’re actually operating. Here’s the whole platform, demystified.
Connect Live to the Warehouse
Connects live to Snowflake, Databricks, BigQuery, Redshift, Synapse, Starburst or Oracle, with no extracts, so the warehouse stays the source of truth.
Model as Code (TML)
Define metrics, relationships and business logic as code in ThoughtSpot Modeling Language (TML), then review, branch, diff and promote models in Git.
AI-Assisted Modeling
SpotterModel builds and enriches the model without code, suggesting descriptions, relationships and synonyms, so it keeps pace with the business.
Govern the Semantic Layer
The model defines certified metrics and relationships once, so ‘revenue’ means the same everywhere — the governed layer that keeps Spotter accurate.
Powers Governed AI Answers
The semantic model grounds Spotter’s AI answers, and Spotter is only as good as that model. TML and live query keep it code-first, current and trusted.
All signals and stores seen together, so each question is answered on the spot, not patched together across silos.
Twelve capabilities. Connect, model, govern.
Modeling & Connectivity is the code-first, governed semantic foundation — TML + live query — that makes AI answers trustworthy, from the ThoughtSpot portfolio.
Live Query on the Cloud Warehouse
Queries your cloud data warehouse live, in place, with no extracts or copies, so answers reflect current data at scale.
Snowflake, Databricks, BigQuery & More
Connects to Snowflake, Databricks, Google BigQuery, Amazon Redshift, Azure Synapse, Starburst and Oracle.
No Extracts, No Stale Copies
Live query means no extracts to schedule and no stale copies to govern, so fewer security and freshness headaches.
Model as Code (TML)
Define metrics, relationships and logic as code in ThoughtSpot Modeling Language (TML), one governed definition everywhere.
Version-Controlled, Git-Friendly
TML is Git-friendly: review, branch, diff and promote data models like software, with full change history.
SpotterModel — AI-Assisted Modeling
SpotterModel suggests descriptions, relationships and synonyms, building and enriching the model without code.
Worksheets, Views & Joins
Model worksheets, views, joins and relationships over warehouse tables into a semantic layer people and Spotter can query.
Certified Metrics & Business Logic
Define metrics, formulas and logic once, centrally, so ‘revenue’, ‘churn’ and ‘active user’ mean the same everywhere.
One Governed Source of Truth
Every answer, dashboard and AI response draws on the same certified definitions, not conflicting spreadsheets.
Row-Level Security & Access
Row-level security and access controls in the model keep self-service safe, so each user sees only their data.
The Grounding for Spotter's AI
This governed model grounds Spotter’s answers, so the AI analyst is accurate and trusted, not hallucinating.
DevOps for Your Data Models
TML in Git brings CI/CD-style workflows to data models: dev/test/prod promotion, peer review and rollback.
Watch ThoughtSpot modeling in action
The overview, getting started, and the core workflows.
SpotterModel, SpotterViz & SpotterCode — the BI A-Team
SpotterModel — AI-assisted modeling.
Introducing the ThoughtSpot Agentic Analytics Platform
The governed foundation, in context.
Create Liveboards
Model-backed analytics in action.
Want a live, India-context walkthrough for your environment?
Book a guided demo →Data is everywhere. Answers, in one place.
Here’s what genuinely sets ThoughtSpot’s modeling layer apart (and where a rival leads).
The semantic model is where AI analytics succeeds or fails — this is that foundation
An AI analyst on raw tables doesn’t know what ‘revenue’ means or how tables relate, so it gives confident wrong answers. The fix is a governed semantic model — worksheets, joins, certified metrics, row-level security — and that model is where AI-analytics projects succeed or fail.
TML — code-first, Git-friendly, version-controlled semantic modeling
Click-built models have no history, review or promotion path. TML defines metrics, relationships and logic as human-readable code in Git, so models are reviewed, diffed, promoted from dev to prod and rolled back, as Looker’s LookML pioneered — and here it directly powers Spotter.
Live query on the cloud warehouse — no data movement, current answers
Extract-based BI creates stale copies and governance gaps. ThoughtSpot queries Snowflake, Databricks, BigQuery, Redshift, Synapse, Starburst and Oracle in place, so answers are current, the warehouse stays the source of truth, and it scales to billions of rows. It suits on-prem or legacy-only shops less well.
One governed source of truth — the same metric everywhere
Without a governed layer, each team defines metrics its own way and meetings become arguments over whose figure is right. The model defines certified metrics and business logic once, with row-level security, so every dashboard, person and AI answer uses the same definition and sees only permitted data.
An India-origin analytics leader — and TechBag adds the local layer
ThoughtSpot was co-founded by Indian-origin engineers Ajeet Singh and Amit Prakash, with R&D in Bangalore since 2017, Trivandrum and Hyderabad. TML suits Indian data teams’ engineering culture on Snowflake and Databricks. TechBag adds semantic-model help, comparison with Looker, dbt, Cube and AtScale, and INR/GST.
The honest scope
Modeling & Connectivity offers TML, live query across many warehouses and SpotterModel, tied to Spotter. Looker’s LookML is the benchmark, dbt Semantic Layer and Cube suit a headless layer feeding many tools, and AtScale leads OLAP scale. It is premium, cloud-warehouse oriented, and TML has a learning curve.
The numbers behind the platform
What your ThoughtSpot journey looks like
Scoping (& capabilities)
Which capabilities and which cloud warehouse? TechBag scopes it and compares Looker, dbt, Cube and AtScale honestly.
Connect & model
Connect live to your warehouse and build the governed semantic model in TML, with SpotterModel helping enrich it.
Govern & version-control
Certify metrics, set row-level security, and keep TML models in Git to review, diff and promote them DevOps-style.
Ground AI & refine
The governed model grounds Spotter’s answers; refine it as the business changes, with SpotterModel. TechBag supports you (GST).
The review scoreboard
Modelled on Gartner Peer Insights structure. *Counts and breakdowns are illustrative pending verified review collection.
Quick poll — what’s driving your evaluation?
Talk to an advisor“Once our semantic model was code in Git — reviewed, diffed, promoted — it stopped being fragile. TML brought real discipline.”
“Live on Snowflake with no extracts, the model always queries current data. No data movement was the deciding factor.”
“We learned the semantic model is where these projects succeed or fail. TechBag helped us build it properly, and Spotter shines.”
“Defining certified metrics once ended the ‘whose number is right?’ meetings, and every AI answer uses the same definitions.”
“As a dbt shop we compared LookML and dbt’s Semantic Layer, but for a model tied to a governed AI analyst, TML won.”
“SpotterModel suggested descriptions, relationships and synonyms, so enriching the model wasn’t all manual.”
“Row-level security in the model means each user sees only their data, which let us open analytics to the business.”
“ThoughtSpot’s Bangalore R&D gave us confidence. TechBag scoped the capabilities, helped build the model and handled GST.”
Where everyone sits — the grids
Analyst firms bury this view behind paywalls, and review-site grids score satisfaction, not fit for your estate. So here’s TechBag’s synthesis of the Semantic-layer & modeling market — tap any vendor to see why it sits where it does.
TechBag Data & Observability Grid
Execution strength vs product vision — the classic market map, minus the paywall.
Code-first model + governed AI. This page.
Model Depth × Governance
The grid nobody publishes — semantic-model depth vs governance.
TML + live-query + governed AI.
Positions are TechBag’s illustrative synthesis of public review-platform data and vendor documentation — not a reproduction of any analyst graphic. Verify before relying on it.
ThoughtSpot (TML) vs the semantic-layer / modeling field
Looker (LookML), dbt Semantic Layer, Cube, AtScale and warehouse-native metrics — honest lanes; the edge is a CODE-FIRST, version-controlled model (TML) tied directly to governed AI analytics and live-query. Benchmark modeling? Looker. Headless layer? dbt/Cube. OLAP? AtScale. We say so.
| Dimension | ThoughtSpot (TML) | Looker (LookML) | dbt Semantic Layer | Cube | AtScale | Warehouse-native |
|---|---|---|---|---|---|---|
| Position | Code-first model + governed AI | Benchmark modeling (LookML) | Modern semantic layer (MetricFlow) | Headless semantic layer | OLAP semantic layer | Native metrics (Snowflake/DBX) |
| Code-first / version-controlled | TML — Git-friendly | LookML — the benchmark | YAML in dbt, in Git | Code-first (JS/YAML) | Model-driven (some code) | Config in warehouse |
| Live query / no data movement | Live on the warehouse | In-warehouse (LookML) | Warehouse-native (dbt) | Queries the warehouse | Live or aggregate cache | Native to the warehouse |
| Warehouse coverage | Snowflake/DBX/BQ/Redshift/Synapse+ | Broad (GCP-centric) | Major warehouses | Major warehouses | Enterprise DWs | Its own warehouse only |
| Integrated with AI analyst | Grounds Spotter directly | Some AI (Google) | Headless — BI-agnostic | Headless — BI-agnostic | Feeds BI tools | Cortex/Genie native NL |
| Headless / feeds many tools | Tied to ThoughtSpot stack | Mostly Looker | Headless — feeds many | Headless — feeds many | Feeds many BI tools | Native to one warehouse |
| Governance / source of truth | Certified metrics + RLS | LookML governance | Metrics as code | Central metrics | Enterprise governance | Warehouse governance |
| Best fit | Code-first model + governed AI on the modern stack | Benchmark code-first modeling | Headless semantic layer for dbt shops | Headless layer feeding many tools | OLAP-scale aggregation | Good-enough native metrics in one DW |
Which data, analytics & observability approach fits you?
Honest fit signals — because the fastest way to lose your trust is to pretend one product wins every scenario.
Choose ThoughtSpot Modeling & Connectivity if…
- ✓You want a CODE-FIRST, version-controlled semantic model (TML — Git-friendly) with DevOps discipline
- ✓You want the model to directly power a GOVERNED AI analyst (Spotter) — not a headless layer sitting alone
- ✓You want live query on your cloud warehouse (Snowflake/Databricks/BigQuery+) — no data movement, current answers
- ✓You're on the modern data stack — with TechBag adding scoping, critical model-building help & GST
Looker (LookML) if…
- ✓You want the benchmark code-first modeling language and deep Google Cloud governance
dbt Semantic Layer if…
- ✓You're a dbt shop and want metrics-as-code / MetricFlow feeding many downstream tools
Cube if…
- ✓You want a headless, BI-agnostic semantic layer feeding many applications and tools
AtScale if…
- ✓You need an OLAP semantic layer with high-performance aggregation at enterprise scale
What could this save you?
Drag the sliders (data models; hour cost as loaded rate). Estimates contrast click-built, ungoverned modeling (no version history, conflicting metrics, stale extracts) vs ThoughtSpot Modeling & Connectivity (code-first TML, one governed source of truth, live query) — the wins are trusted AI answers, less rework, and a version-controlled model. NB: premium pricing — TechBag scopes your case. Illustrative.
Loaded cost = salary + overheads per productive hour. Illustrative only — your TechBag quote models your actual environment and modules.
Three ways to consume it
ThoughtSpot is PREMIUM and quote-priced (editions have included Pro and Enterprise; pricing trends toward consumption/credits for AI workloads), in USD — no simple public list. Enterprise deployments commonly land well into six figures (third-party estimates cite ~$137K/yr average; treat as indicative). Success depends on a well-built semantic model — where these projects succeed or fail. TechBag scopes the capabilities you need, helps build the model, and quotes current figures with INR/GST.
ThoughtSpot (premium, by quote)
Best for a governed model powering AI analytics
- Modeling & Connectivity: TML (code-first) + live query + SpotterModel
- Live on your cloud warehouse (Snowflake/Databricks/BigQuery+) — no data movement
- Editions Pro/Enterprise; consumption trend for AI — quote-based, USD
+ Platform add-ons
Best for a broader rollout
- Scoped to your estate
- Add-on modules as needed
- Phased, right-sized deployment
+ semantic model & local support
Best value with TechBag
- Critical semantic-model help (worksheets, TML, certified metrics — where projects succeed or fail)
- Honest Looker / dbt Semantic Layer / Cube / AtScale comparison
- TechBag adds onboarding, INR/GST invoicing & local support
Buy it for less — TechBag pricing beats list
Whatever the list prices above, TechBag negotiates a significantly better deal — with GST-compliant INR invoicing and local support. Ask us for your discounted quote.
Get an India-ready quote
Tell us your requirements and current tools — we’ll model it against what you spend today.
The 8 questions to ask every vendor
Take this into your next vendor call — including ours.
Who will build the governed model (TML)? It is where AI-analytics projects succeed or fail.
Want version-controlled models? TML is Git-friendly — review, branch, diff and promote data models DevOps-style, like software.
On Snowflake/Databricks/BigQuery? ThoughtSpot connects live — no extracts, no data movement, current answers at warehouse scale.
Want to move faster? SpotterModel builds and enriches the model — suggesting descriptions, relationships and synonyms.
Tired of arguing whose number is right? Can you certify metrics once, centrally, with row-level security?
The model is what grounds Spotter's answers — governed, not guessed. Get the model right and the AI is trusted.
Weighing Looker (LookML), dbt Semantic Layer, Cube or AtScale? TechBag compares honestly for your stack.
ThoughtSpot is India-origin with huge Bangalore R&D — TechBag adds local scoping, model help, INR/GST and support.
Questions buyers ask
Ready to build the governed foundation that makes AI answers trustworthy?
Scope ThoughtSpot Modeling & Connectivity (code-first semantic modeling with TML — Git-friendly, version-controlled — plus live query on your cloud warehouse) — and let a TechBag advisor scope the capabilities, help build the semantic model (where success lives), compare vs Looker/dbt/Cube/AtScale honestly, and add INR/GST and local support.
Stats, ratings, review counts and pricing are illustrative and sourced from public materials; verify before purchase.