Reflection AI

https://reflection.ai/

Founded 2024 by Misha Laskin and Ioannis Antonoglou (both ex-DeepMind, AlphaGo / Gemini RLHF). Building autonomous coding agents — Asimov is their code-research agent. Raised $130M Series A March-2025 at $555M valuation led by Lightspeed + CRV. Positioned as Devin/Cognition competitor with RL / robotics-research pedigree.

At a glance

Type
Autonomous coding-agent foundation lab (Misha Laskin / Ioannis Antonoglou)
Tier
T2
Created
2024 (founded)
Latest release
not applicable — orchestration platform, not memory product
License
not applicable — not OSS
GitHub
not applicable — no GitHub repo
Pricing
Private beta; enterprise sales early
Funding
$130M Series A Mar-2025 ($555M val; Lightspeed + CRV; Sequoia, Nvidia investors)

Taxonomy

storage
none-trivial
retrieval
none
persistence
session-only
update
agent-controlled
unit
turn
governance
opaque
conflict
out-of-scope

When to use

Optimised for: autonomous coding research agents; long-horizon software-engineering tasks

Anti-fit: not applicable — orchestration platform, not memory product

Pros & cons

Pros

Strongest ex-DeepMind RL pedigree in coding-agent space; $130M at $555M val is largest among 2024-vintage coding-agent labs; differentiated by autonomous research focus (Asimov).

Cons

Pre-product launch (private beta 2025); Devin / Cursor / Cognition / Anthropic Claude Code already established; small team vs commercial peers.

Claims & capabilities

$130M Series A Mar-2025 at $555M valuation (Lightspeed + CRV); founders ex-DeepMind AlphaGo + Gemini RLHF team; Asimov code-research agent product

Technical surface

API surface
not applicable — orchestration platform, not memory product
Backend storage
not applicable — orchestration platform, not memory product
Deployment
Managed cloud (private beta)
Embedding model
not applicable — not a memory product
Multi-tenancy
not applicable — orchestration platform, not memory product
MCP
not applicable — orchestration platform, not memory product
A2A
not applicable — orchestration platform, not memory product
OpenTelemetry
not applicable — orchestration platform, not memory product

Similar systems

Other multi-agent orchestration platforms in the catalog, ranked by inbound references.

  • Adept ACT T3

    Founded 2022 by ex-OpenAI / Google researchers (David Luan, Kelsey Schroeder). Built ACT-1 / ACT-2 multimodal action transformers for computer-use. **Acquired by Amazon June-2024 (acqui-hire)** — co-founders + key team joined Amazon AGI; Adept the company continues with Zach Brock as remaining executive. Important historical entry — ACT models inspired Anthropic Computer Use + OpenAI Operator.

  • Burr (DAGWorks) T2

    Burr is a state-machine framework for LLM agents from DAGWorks (commercial company behind Hamilton dataflow framework, $4M seed). State-machine abstraction instead of DAG / ReAct — more debuggable for production. Targets the same niche as LangGraph but with state-machine semantics rather than DAG. BSD-3 license. Founded 2023 by ex-Stitch Fix / Two Sigma / Lyft engineers.

  • CrewAI Enterprise T2

    Commercial Enterprise tier of the CrewAI OSS multi-agent framework. Adds managed cloud runtime, RBAC, audit logging, SLA support, on-prem option. OSS CrewAI is in the Framework-embedded memory section already; this row covers the commercial cloud platform separately. $18M Series A Oct-2024 (Insight Partners). 30k+ companies have deployed CrewAI agents in prod (vendor-stated).

  • Imbue (formerly Generally Intelligent) T3

    Founded 2017 as Generally Intelligent; rebranded Imbue 2023. Building foundation models for reasoning agents that write/edit code. Founders Kanjun Qiu + Josh Albrecht. $200M Series B 2023 ($1B valuation, Astera + Nvidia + Notion's Akshay Kothari). Trained 70B model for reasoning. Most public output is research blog + 'Carbon' OSS coding agent. Pulled back commercial roadmap post-2024.

  • InstructLab (Red Hat / IBM) T2

    Red Hat / IBM-stewarded open-source framework for community-driven LLM alignment + fine-tuning. Method paper 'LAB: Large-Scale Alignment for Chatbots' (Sudalairaj et al, IBM Research 2024). Built on the Granite model family (IBM's open-weights). Apache 2.0. Used in IBM watsonx + Red Hat OpenShift AI for fine-tuning workflows. Not strictly a multi-agent framework — included as 'orchestration platform' because of its role as a substrate for building domain-specific fine-tuned agent models.

  • Lindy T2

    Lindy is a no-code multi-agent automation platform from Florent Crivello (ex-Teleport, ex-Twitter). Targets consumer / SMB segment with email triage, meeting scheduling, CRM-update agents. $50M Series A Oct-2024 led by Andreessen Horowitz + Sequoia. Distinct from Zapier / n8n traditional automation by putting LLMs at the core of every workflow node.

Row last verified 2026-05-14. Catalog data is CC-BY-4.0 — see how to read this.