Covariant

Explainers

Plain-language answers to the questions people ask about Covariant, from DEPLOY’s explainer library. Each is written in the language of the question and cross-checked against this registry.

• What is Covariant RFM-1?

  Covariant RFM-1 (Robotics Foundation Model 1) is Covariant's foundation-model-for-robotic-manipulation, announced March 2024. CRITICAL DISTINCTION per Agent A source-depth hygiene: corporate-state vs model-state separation. Corporate-state: Covariant founders + ~25% staff Amazon-acqui-hired August 2024 (Pieter Abbeel + Peter Chen + Rocky Duan + others moved to Amazon AGI); Covariant continues as IP/legal entity. Model-state: RFM-1 has NO verified production deployment; the verified-deployment Covariant model is the legacy Brain (deployed via KNAPP warehouse-automation partnership), NOT RFM-1. Cohort positioning: foundation-model-for-robotic-manipulation tier distinct from foundation-model-for-robotics (PI) + foundation-model-for-driving (Wayve) + foundation-model-for-physical-AI-general (NVIDIA GR00T) + foundation-model-for-dexterous-manipulation (Dyna) cohort tier positions.

• How did DEPLOY correct the Covariant corporate-state narrative?

  Aggregator coverage of [Covariant](/explainers/what-is-covariant-rfm-1) frequently asserts the company was acquired by Intrinsic (Alphabet's industrial robotics arm). Per Agent A primary-source verification, the framing is wrong. Covariant operates continuing-but-diminished post-Amazon-AGI departure August 2024. Founders Pieter Abbeel + Peter Chen + Rocky Duan + ~25% of staff moved to Amazon's AGI team; Covariant the legal entity continues operating with Ted Stinson as CEO; legacy Covariant Brain continues verified commercial deployment via the KNAPP warehouse-automation partnership; non-exclusive license and IP retained by Covariant. The within-corporate-state framing is editorially distinctive: Covariant is simultaneously verified-continuing (corporate entity + CEO + commercial deployment via KNAPP) and verified-diminished (founder + leadership + capacity moved to Amazon AGI; reduced staff; research-tier RFM-1 model future cap-flagged). Two true claims at once; aggregator framings collapse them into a single wrong claim ('Covariant acquired by Intrinsic' or 'Covariant defunct after Amazon acqui-hire'). This piece documents the catch as framework-in-action worked example: corporate-state-precision discipline + within-entity verified-continuing-and-verified-diminished framing + acquirer-identification verification operating at editorial-anchor depth.

• How does DEPLOY track acquisition history state?

  DEPLOY tracks acquisition history state as a five-structure taxonomy operating at relationship-graph granularity per Agent A's Arc D substrate (29 acquisitions + 20 FK-acquirer + 9 external + 7 acquired-assets typed). The structure taxonomy: full_acquisition (acquirer absorbs target entirely; target ceases as independent entity) → asset_purchase (specific assets transfer; target may continue) → acqui_hire (team transfers; assets minimal) → license_and_hire (Amazon × Covariant canonical: RFM models licensed + team transitions; Covariant remains standalone under Stinson; ~25% staff transition documented) → spac_merger (de-SPAC pattern; Sarcos → Palladyne canonical example). The four-state valuation_basis discipline: exact (SEC-disclosed or court-record) versus reported (press-release; not SEC-verifiable) versus undisclosed (transaction confirmed but valuation not disclosed) versus contingent (ZB × Monogram CVR canonical: structured contingent payment with earnout). The canonical lesson banked at validator-discipline depth: exact name match only, never alias-contains, in M&A graphs (6 acquisitions reverted from alias-contains reconciliation bug). Cap-flag-as-trust-signal operates recursively on acquisition framing.

• How does DEPLOY track cross-cluster talent-flow as diaspora graph?

  DEPLOY tracks cross-cluster talent-flow at primary-source-anchored PersonCompany-edge granularity per Arc A people graph substrate. The diaspora graph framework operates at three canonical pattern classes: post-wind-down diaspora (Cruise canonical worked example; founders + executives + technical leadership transition across multiple destinations after corporate wind-down); license-and-hire diaspora (Amazon × Covariant canonical worked example; co-founders + ~25% staff transition to acquirer while standalone entity continues under remaining leadership); adjacent-employer-prior diaspora (Meta AI / FAIR + Google X / Everyday Robots as recurring prior employers in brain-providers and humanoid cohort hires). Each pattern class operates at distinct PersonCompany-edge structure: current_role=false with end_date populated + where-they-went edge (post-wind-down + license-and-hire); current_role=true with prior-employer edge at honest-absence end_date if no specific tenure-end disclosed (adjacent-employer-prior). The framework reads talent-flow at four substrate-axis granularity: source company + destination company + role transition + tenure date precision; cap-flag-as-trust-signal operates recursively on diaspora framing same as on any other relationship-record claim depth. Cross-property bidirectional discipline operational: PersonCompany edges cross-reference acquisition records (Cruise wind-down ↔ GM full re-absorption Acquisition record; Covariant license_and_hire ↔ Amazon × Covariant Acquisition record) + partnership records + entity records simultaneously.

Current leadership (2)

• Tianhao Zhang Co-founder (current lead)secondary-verified
• Ted Stinson CEOsecondary-verified

Founders (4)

• Pieter Abbeelcofounderfounded 2017-01-01no longer at company
• Peter Chencofounderfounded 2017-01-01no longer at company
• Rocky Duancofounderfounded 2017-01-01no longer at company
• Tianhao Zhangcofounderfounded 2017-01-01

Former / Previously (3)

• Pieter Abbeel Co-foundersecondary-verified
• Peter Chen Co-founder & CEOsecondary-verified
• Rocky Duan Co-founder & CTOsecondary-verified

Safety record

No incidents on record for Covariant.

Only active incidents are counted. Retracted incidents are excluded from this summary but remain reachable at their canonical URLs.

Brains developed (1)

• RFM-1foundation-model · research

Financial state

Reporting basis

aggregator_estimate

Lifecycle stage

wound_down

Counterparty risk class

elevated

Each numeric field carries its own basis marker. Aggregators report a number; this surface preserves the source class so verification depth travels with the value.

Funding rounds (2)

• Series C2021-07-27

  $80M(reported)

  Investors: Index Ventures (lead), Amplify Partners, Radical Ventures, Canada Pension Plan Investment Board (CPP Investments), Temasek

• Series B2020-05-01

  $40M(reported)

  Investors: Amplify Partners (lead), Index Ventures, Radical Ventures

Acquisitions (1)

• acquired by Amazonlicense and hire

Recent activity

Every change to this record is dated, sourced, and independently verified where marked.

• Verified media added✓VerifiedJun 11, 2026

  RFM-1: A world model that understands physics

• Verified media added✓VerifiedJun 11, 2026

  Covariant Robotic Induction | Explainer Video

• Record createdJun 3, 2026

  Added to the verified registry

Full change history →

Sources (3)

1. https://www.businesswire.com/news/home/20240311948570/en/Covariant-Introduces-RFM-1-to-Give-Robots-the-Human-like-Ability-to-Reason ↗
2. https://www.aboutamazon.com/news/company-news/amazon-covariant-ai-robots ↗
3. https://www.geekwire.com/2024/amazon-hires-covariant-founders-inks-licensing-deal-with-robotics-ai-startup-in-latest-reverse-acquihire-deal/ ↗

Common questions

What is Covariant?

Robotics-AI company (Berkeley area; founded 2017) behind RFM-1, a robotics foundation model that powered warehouse pick-and-place robots; largely absorbed by Amazon (reverse-acquihire, Aug 2024).

What does Covariant make?

Covariant is a robot brain developer: 1 foundation model or autonomy stack on the DEPLOY registry are credited to Covariant.

Is Covariant publicly traded?

Covariant is not an independent public company; it has been acquired, per the DEPLOY registry.

Who is the CEO of Covariant?

Tianhao Zhang is the ceo of Covariant, per the DEPLOY registry.

How can I invest in Covariant?

Covariant has no separately traded stock; it has been acquired.