Atomistic Insights uses physics-based AI to reveal how proteins move and uncover hidden binding sites, enabling faster discovery of more effective medicines.
Atomistic Insights is an AI-driven drug discovery platform that makes previously intractable targets druggable by addressing a core limitation in small-molecule discovery: reliance on static protein structures.
Most tools miss transient and cryptic pockets that only appear during protein motion—the sites where many selective and potent drugs bind. Traditional molecular dynamics captures motion but is too slow for routine use, while leading LLM-based tools like AlphaFold 3 excel at static predictions but lack native dynamics modeling due to limited high-quality training data.
Our platform, DeepPath, uses physics-guided AI to predict realistic protein motion and generate compact conformational pathways that expose hidden binding opportunities. We deliver high-quality protein dynamics in hours—versus days or weeks with conventional methods—providing actionable ensembles that help teams prioritize targets, design smarter experiments, and accelerate campaigns.
Currently applied with biotech partners on inflammation and antiviral targets, DeepPath is evolving to cover small-molecule binding, protein-protein interactions, molecular glues, and PROTAC-like modalities. With tight feedback loops to partner assays, Atomistic Insights is positioned to transform early-stage drug discovery.
Leadership Team
- Katie Kuo, PhD – CEO
Leads discovery and partner programs. Co-inventor of DeepPath with proven track record running collaborations that generated prospective binders and mechanism data. - Professor JC Gumbart – CTO
Renowned biophysicist known for bridging physics and biology. Guides physics-informed architecture and rigorous benchmarking against NMR, HDX, and cryo-EM data. - Shiyao Bao – CBO
Strategy and pricing expert who has taught revenue analytics and conducted over 150 customer interviews (NSF I-Corps and accelerators). Drives go-to-market strategy and partner design. - Andrew Pang – Scientific Advisor
Supports computational pipeline design and validation.
The team combines deep scientific expertise, enterprise selling experience, and strong product execution. 50% female-founded and globally trained, with a filed US utility patent, secured venture/grant support, and early partnerships validating both technology and delivery capabilities.
Atomistic Insights lands customers through targeted paid pilots that convert into deeper co-discovery partnerships, while originating select internal programs for future licensing or sale.
Commercial Strategy
- Pilot-to-Co-Discovery Model
Paid pilots on partner-defined targets with clear success criteria (pocket triage accuracy, NMR/HDX correlation, time-to-hits). Successful pilots transition to co-discovery agreements with milestone economics and options to license/acquire assets. - Internal Asset Origination
One to two in-house programs matured to mechanism data and lead series for partnering or sale. - Platform Access
Workflow-level integration of DeepPath dynamics to accelerate partner campaigns.
Current & Developing Partnerships
- Synovel Laboratory, DevsHealth, Avalon Biosciences
- Academic collaboration with Professor Thomas Barker (University of Virginia)
- Astellas Pharma collaboration in finalization
- Seeking additional pharma partners (inflammation, antiviral, transcription factors), biophysics labs (NMR/HDX/cryo-EM), CROs, compute partners, and disease foundations (e.g., choroideremia)
Founder-led business development targets VPs of Chemistry, Heads of Research, and platform leaders via warm introductions, technical seminars, and benchmark case studies.
Atomistic Insights generates revenue across three reinforcing tracks designed for near-term cash flow and long-term upside.
Revenue Tracks
- Pilots & Platform Access
Paid pilots convert to annual platform subscriptions (per-target or per-program pricing) with usage-based expansion options—delivering immediate revenue and sticky collaboration. - Co-Discovery Economics
Upfront payments, study-based milestones (hit confirmation, mechanism data), success milestones (lead series, IND-enabling), and low single-digit royalties on approved products. - Asset Transactions
Out-licensing or sale of internal/co-discovered programs with upfronts, clinical/commercial milestones, royalties, or cash/equity deals.
Near-term revenue from pilots/platform access; mid-term dominated by co-discovery milestones; longer-term upside from asset transactions. Pricing is budget-friendly at pilot stage while rewarding success at later gates.
Investment Opportunity
Atomistic Insights is raising $1 million to expand the team, advance DeepPath, deepen partnerships, and initiate the first in-house discovery program—providing ~18 months of runway to key value-inflection milestones.
Use of Proceeds
| Allocation | Percentage | Purpose |
|---|---|---|
| Team Expansion | 40% | Hire 2–3 key roles in physics-informed AI, computational chemistry, and program management. |
| Platform Development | 25% | Broader motion classes, improved active learning, hardened workflows for reproducibility, auditability, and secure deployment. |
| Partner Validation | 20% | Joint studies with biotech/pharma, NMR/HDX/cryo-EM benchmarks, limited CRO assays. |
| Internal Discovery Seed | 10% | Target nomination, virtual screening on DeepPath ensembles, hit triage with partner assays. |
| Operations & IP | 5% | Cloud compute, compliance, and patent work. |
Key Milestones (End of Runway)
- Convert ≥2 pilots to co-discovery agreements with upfronts and milestones
- Demonstrate prospective binders confirmed by assays on ≥2 targets
- Release faster, more validated platform version
- Nominate first internal asset with validated hit series and mechanism data
DeepPath integrates physics-based simulation with AI to generate realistic protein conformations without requiring massive pre-existing datasets.
Completed integration of diffusion model with strong retrospective validation (accurate potency ranking) and prospective binder identification. Results have driven partnerships with Synovel Laboratory, DevsHealth, Avalon Biosciences, Professor Thomas Barker (UVA), and Astellas Pharma (in finalization).
Current Capabilities
- Predicts dynamics from single structure or endpoint pairs
- Physics-informed active learning captures transient conformations missing from PDB
- Delivers compact pathways and ensembles in hours for pocket triage and mechanism hypotheses
Validation & Readiness
- Retrospective benchmark success
- Prospective binders identified
- NSF I-Corps (100+ interviews) and Alchemist Accelerator graduation
Immediate Next Steps
- Extend to protein–nucleic acid systems (transcription factors, oncology)
- Apply to choroideremia (novel molecular glue therapeutics)
- Expand experimental benchmarking (NMR, HDX, cryo-EM)
Atomistic Insights has secured the following non-dilutive grants and early venture investment:
| Source | Amount | Type |
|---|---|---|
| I-Corps Sites, Georgia Tech | $3,000 | Grant |
| VentureLab, Georgia Tech | $10,000 | Grant |
| Georgia Research Alliance Phase IA & IB | $50,000 | Grant |
| NSF I-Corps Program | $50,000 | Grant |
| TechReady Grant, Georgia Tech | $25,000 | Grant |
| Alchemist Accelerator SAFE | $54,000 | Venture Investment ($10M post-money cap) |
