Press kit.

Apik Systems is a frontier-research lab structured as a company. The distinction matters: a research-driven company has the capital, the engineering depth, and the deployment surface that pure academic labs do not, and Apik uses that to run research programs that require long-horizon engineering investment. The company shape is in service of the research, not the other way around. The closest peer in the safety-first framing is Anthropic; the closest peer in research breadth is DeepMind; the notable departure from OpenAI is on the question of how aggressively a frontier lab should commercialize.

Apik is safety-first in a specific operational sense: the same Responsible Development Policy that gates research releases also gates product releases, and the disclosure schedule is identical for both. This is stricter than the consumer-AI norm, which typically maintains a research-side safety policy alongside a separate and looser product-release process. Apik does not run that separation. The cost is occasional product velocity; the gain is that the policy actually binds at the surface where users meet the system.

Apik is open-by-default in the sense that the default disposition of research output is publication — into peer-reviewed venues, into the news surface on this site, and into engineering writeups that are detailed enough for external researchers to engage with. The exception clause for held-back research is narrow and is documented in the transparency framework. Where we hold work internal, we disclose the existence of the held-back artifact even when the artifact itself stays internal. This is the canonical three-paragraph framing media should use; longer pieces should source the framing back to the manifesto, the Responsible Development Policy, and the transparency framework.

Shailesh leads Apik Systems as CEO, with responsibility for the lab's research agenda, partnerships, and operational direction. He sets strategy across the eight research pillars and three internal projects, and represents Apik publicly on the questions where the institutional shape of the lab is doing as much work as the technical research: AI safety as a precondition for deployment, abundance economics as the macroeconomic problem the company is structurally betting on, and the institutional architecture required to deploy autonomous systems without concentrating coordination authority.

His focus sits at the boundary between research direction and deployment. The recurring question in his work is how a small lab translates long-horizon technical bets into the partnerships, capital structure, and policy posture that keep them open-by-default and verifiable — which mechanisms preserve the research-grade discipline of the lab as the lab scales, and which ones erode it. He works closely with the founding team on the manifesto, the Responsible Development Policy, and the transparency framework, and sits on the safety council that gates external releases.

For background, Shailesh's LinkedIn is at linkedin.com/in/shailesh-tamgadge. Press inquiries that require him specifically should be routed via press@apiksystems.com with the topic and deadline noted.

Rehan co-founded Apik Systems and serves as CTO. He leads research and engineering across the eight pillars and three internal projects — owning the technical roadmap, the experimental rigs, the evaluation infrastructure, and the deployment systems on which the lab's work depends. The unifying job is keeping research direction and engineering reality coupled, so that the claims the lab publishes survive contact with the systems that eventually ship and the policies the lab commits to remain operationally binding rather than aspirational.

He is also the principal author of Apik's manifesto, safety principles, and Responsible Development Policy, and the lead author on the company's open-by-default research programme. His research interests cluster around four overlapping problems: coordination as a computational problem; mechanism design for hybrid human-AI economies; verified envelopes around learned policies (the technical line on which Project Aegis sits); and embodiment as a closing condition for safe autonomy.

He is the lab's primary spokesperson on AI safety, abundance economics, and the technical questions that connect the two — and is the point of contact for journalists writing on the agentic-systems and interpretability sides of the research.

For background, Rehan's LinkedIn is at linkedin.com/in/rehan-temkar-94a418105 and his X handle is @RehanTemkar. Direct contact for technical-press inquiries: rehan@apiksystems.com.

Both co-founders speak on behalf of the lab. The split below reflects the domain on which each is the strongest available source.

• Shailesh Tamgadge
  Company strategy, partnerships, abundance economics, governance, the institutional architecture of frontier AI, the lab's posture on policy questions.
• Rehan Temkar
  Technical research direction, AI safety, the manifesto, agentic systems, interpretability, formal-methods work, the Project Aegis line.

Both are available for on-record interviews on a schedule honest about availability — book through press@apiksystems.com with a topic, deadline, and outlet. We try to respond within two business days. We do not pre-clear quotes, but we do provide context, source links, and (where appropriate) a draft response in writing for fact-checking. For background-only conversations, name the constraint at the start of the booking.

• How big is Apik? A small founders-and-research-fellows team, with associated software agents and embodied platforms. We hire patiently and at depth; the company is growing slowly on purpose. We do not currently disclose headcount.
• Are you a startup or a lab? Both. Apik is structured as a research-driven company; the research is the deliverable. The company shape exists to fund and protect long-horizon research, not the other way around.
• What do you sell? Three products today — Senwitt, Brello AI, and Surfacedd — that ship under our Responsible Development Policy. Senwitt is a cognitive infrastructure product; Brello is a generative intelligence product; Surfacedd is a substrate for agent-mediated commerce. The products are public surfaces of the civilization-stack research; the long-horizon work is the research itself.
• Are you safety-first? Yes, in a specific operational sense: the Responsible Development Policy gates research and product releases identically, and the transparency framework binds both. This is stricter than the consumer-AI norm and is documented at /safety/responsible-development-policy and /safety/transparency.
• Are you funded? Yes; we do not currently disclose funding terms. Funding-related questions go via partnerships@apiksystems.com.
• What's the relationship to Anthropic / OpenAI / DeepMind? No formal relationship. The lab is in dialogue with the published practice of all three on safety frameworks and transparency norms; the manifesto and the Responsible Development Policy both name those engagements directly. We are not a wrapper, an integrator, or a downstream consumer of any single frontier provider.
• Where are you based? Remote-first with periodic on-site presence; the legal entity is documented in the standard places. We do not disclose the on-site location publicly.

Lab name on first reference: Apik Systems. Subsequent references: Apik. Both spellings without honorifics, without italicization, without the article. The URL of record is apiksystems.com. When citing a publication on this site, use the canonical URL (the page's permalink); titles change occasionally but URLs do not. When citing an Apik person, use the role at the time of the quote (Co-founder & CEO; Co-founder & CTO).

We do not pre-approve quotes; once a quote is on the record it stays on the record. We do provide context, source links, and a written follow-up response for fact-checking on request. Where we are quoted out of context in a published piece, we ask for a correction; the request goes via the outlet's standard correction channel and we publish the correction request alongside any response on the news surface where appropriate.

Image rights: photographs supplied via the press kit are licensed for editorial use in connection with reporting on Apik Systems. Reproduction in advertising, training data, or generative-model corpus building is not licensed.