Ashton Teng
@ashtonteng
Followers
153
Following
82
Media
11
Statuses
94
Co-Founder & CEO @keplogic - building Kepler AI bioinformatics engineer, biotech and longevity enthusiast 🧬🌈💻🧠⛷️
San Francisco, CA
Joined November 2014
This unlocks:.✅ Safe AI code execution.✅ Strict team/project boundaries.✅ Pause & resume workflows.✅ Versioning & revertability. The result: AI agents that can handle the messy realities of life sciences R&D—without compromising trust, compliance, or collaboration.
0
0
1
That’s why we built our platform on open-source micro-VM sandboxes from @e2b . Each agent task runs in its own lightweight, hardware-isolated VM. On top of E2B, we’ve added optimizations for high performance life sciences workloads.
1
0
2
Most teams start by running agents on laptops or shared servers. 👉 Great for prototyping.👉 But cracks show fast: hardware limits, no isolation, risky execution. Next stop? Containers. Useful for packaging, but not isolated enough for R&D.
1
0
1
When an AI agent isn’t just answering questions—but actually writing and executing code on research data—you need:.🔒 Secure code execution.⚡ Massive compute + memory handling.⏸️ Pause & resume for long analyses.
1
0
1
🧱 Building the Right Infrastructure for AI Agents in Life Sciences R&D. Full blog: At Kepler AI, we build enterprise-grade AI agents to accelerate scientific discovery. One key lesson: the infrastructure matters just as much as the AI itself. 🧵.
1
5
17
🚀 GPT-5 is here — and now live in @kepler_ai_ !. With the launch spotlighting medical research use cases, we were eager to see how it performed on our internal benchmarks of real bioinformatics research tasks. We are impressed - here are some highlights on one research task on
0
1
1
Since TahoeDive on @kepler_ai_ launched a month ago we have seen an overwhelming positive response from the community, with hundreds of researchers around the world performing massively parallel analyses on the rich Tahoe-100M perturbation dataset. Results are already going into
1
5
13
RT @quinn_leng: 🧵Why Every Startup Will Be AI-Native — or Die Trying. Over the past two months, we have contributed more than 1300 commit….
0
4
0
RT @quinn_leng: What a wild month of product building. @DevinAI is quickly becoming our company’s top contributor—and its monthly “salary”….
0
4
0
RT @_annhuang: In the wake of all the Bio x AI updates, I'm also here to spread awareness of important cat-related research. Isparta et al….
0
1
0
RT @CorinWagen: The past few years of "AI for life sciences" has been all about the models: AF3, NNPs, PLMs, binder generation, docking, co….
0
32
0
RT @saranormous: The “AI leapfrog effect” is quietly transforming sectors historically slow to adopt technology—legal, healthcare, logistic….
0
36
0
RT @_annhuang: 1/12.Excited to share our team's latest work and the first @xaira_thera preprint! Here, we introduce FiCS Perturb-seq, an in….
biorxiv.org
The rapid expansion of massively parallel sequencing technologies has enabled the development of foundation models to uncover novel biological findings. While these have the potential to significan...
0
16
0
Honored to be a part of this huge team effort to publish the largest publicly available genome-wide Perturb-seq dataset! The folks at Xaira have been the most talented people I've worked with. Can't wait to see what everyone discovers from this amazing dataset.
🚀 Xaira Therapeutics has just dropped a game-changer for AI-driven biology. Today, we unveiled X-Atlas/Orion, the largest publicly available genome-wide Perturb-seq dataset to date—spanning 8.4 million single cells with perturbations across all ~20,000 human protein-coding
0
0
4
2 months ago Kepler was only an idea - today it was surreal to have Kepler in action on stage at the Databricks AI summit. I'm so grateful for my partner @quinn_leng and our visionary partners @iamjohnnyyu @nalidoust at @tahoe_ai - lets continue to ship fast, together!
0
0
3
It's been amazing working with @iamjohnnyyu , @nalidoust and the @tahoe_ai team - once we talked it was clear that our visions were aligned and we had to build something together, and quickly. This is only the beginning, can't wait to see what's ahead of this partnership!.
Today, we @tahoe_ai are announcing TahoeDive, built in partnership with Kepler AI @keplogic + opening beta access!. TahoeDive is an AI agent for biologists to query & analyze our Tahoe-100M dataset using natural language, with added context from broader scientific literature. 🧵
0
0
0
RT @nalidoust: Today, we @tahoe_ai are announcing TahoeDive, built in partnership with Kepler AI @keplogic + opening beta access!. TahoeDiv….
0
19
0
RT @nalidoust: We’re excited to invite a select group of biologists to try it and share feedback. Request beta access, read more, and see….
tahoebio.ai
Tahoe Therapeutics has partnered with Kepler to build TahoeDive. It uses language and reasoning models and makes it easier to interact with the data for cancer and computational biologists.
0
4
0
RT @nalidoust: We tried to “rediscover” that using TahoeDive — and guess what?. It worked. 🧪 Demo here:
app.getkepler.ai
Enterprise-Grade Bioinformatics Agent Accelerating Scientific Discovery
0
2
0
RT @nalidoust: It’s still in beta and not yet general-purpose — but we’ve tested it against real use cases. Example: Last week, we publish….
0
2
0