BIOVIZIONS · PHASE 1

24-Month Roadmap Dashboard

Quarter → Month → Week. Quarters open to the monthly view; expand any month for the weekly breakdown. Month 1 Week 1 links to editable founding & hiring documents (Documents panel below). Goal at M24: 8–12 validated, patentable combinations + the algorithm kept proprietary (trade secret), Seed-ready (~$2.0M raise · ~$4.75M Seed).
LegalTeamOpsDataAI/MLWet-labIPBizBoardGateDecision
Q1Months 1–3Stand up — governance, team & data foundation
Legal★ W1: Founders' Agreement + CIIA/IP assignments; engage corporate (1b) + IP (1c) counsel. W2: issue founder shares + 83(b)
TeamW1: post Senior ML + Ops Coordinator JDs. W2: onboard Ops Coordinator; first platform/data interviews
OpsW1: data-room shell + index, brand/logo start, Google/Anthropic/Nvidia credits. W2: Carta cap table + 10% ESOP + 409A; Claude API wrapper
DataW3–W4: founders (KZ, Prateek, Prasan) push DB curation, analyses, infra & experiment scoping — twice-weekly syncs (Mon & Fri)
BoardW1: build independent-director shortlist; W2: discuss (decision kept open)
Weekly breakdown
W1
  • Draft & circulate the Founders' Agreement — full per-founder roles & responsibilities 📄 open & edit draft
  • Draft & sign CIIA / IP-assignment agreements (all founders) 📄 open & edit draft
  • Engage corporate counsel (1b) + IP counsel (1c)
  • Post the Senior AI/ML Engineer JD 📄 open & edit draft
  • Post the Operations & People Coordinator JD 📄 open & edit draft
  • Stand up the data-room shell (DocSend/Drive) — load the full document index 📄 open & edit draft
  • Brand + logo brief to a designer (runs into W2); file the 'Biovizions' trademark
  • Submit Google / Anthropic / Nvidia compute-credit applications
  • Onboard the 2 Eastern-pharmacology advisors (advisory agreements incl. IP assignment)
  • Founders begin curating ChEMBL / PubChem / DrugComb subsets
  • Build the independent-director (board) shortlist + warm intros
W2
  • Issue founder stock to Prateek, Kourosh & Prasan (RSPAs); file 83(b) within 30 days
  • Cap table + 10% ESOP in Carta; kick off the 409A valuation
  • Set up Claude API + internal wrapper (Zero-Data-Retention; BAA if PHI)
  • Onboard the Operations & People Coordinator; Deel/Rippling payroll + EOR live
  • First-round interviews: platform + data engineers
  • Brand + logo: iterate to v1; apply across site + data room
  • Board: discuss the shortlist — keep the appointment decision open-ended
W3
  • Twice-weekly founder syncs begin — Mondays & Fridays (KZ, Prateek, Prasan) to drive every angle
  • DB curation continues: extend ChEMBL / PubChem / DrugComb; add NCI-ALMANAC subsets
  • Kourosh: model scoping — shortlist base models (ChemBERTa-2 / MoLFormer / Chemprop); sketch Engine A/B architecture + evaluation plan
  • Prateek: draft assay design (384-well checkerboard / FIC); scope collaborator labs; draft data schema (molecules / assays / combinations)
  • Prasan: indication shortlist (lead = UTI / AMR); begin the Dravya-guṇa priors list with the advisors
W4
  • Stand up a lightweight analysis environment (notebooks + RDKit) so founders can query the curated data
  • Kourosh: first exploratory embeddings / representation checks on the curated subset
  • Prateek: confirm 1–2 collaborator labs; written lab agreements (with IP terms) in motion; finalize the assay-protocol draft
  • Prasan: encode the first Ayurveda / TCM priors with the advisors; short indication-rationale memo
  • Board: continue shortlist outreach (decision stays open)
  • Month-end review at the Friday sync — curation, analyses & experiment scoping on track
TeamW5–W6: Platform Engineer + Data Engineer #1 + 2 Assay Scientists onboard (from the M1 interviews)
DataStand up Postgres + RDKit cartridge; ingest + normalize ChEMBL / PubChem / DrugComb / NCI-ALMANAC into clean, versioned training tables
AI/MLFinalize Engine A/B architecture + base models (ChemBERTa-2 / MoLFormer / Chemprop) — Kourosh + founders (senior ML not yet in)
OpsHU/IN contracts live via EOR; bookkeeping live (Pilot/QuickBooks)
◆ DECISION: assay-scientist hire timing ↗
Weekly breakdown
W5
  • Platform Engineer onboards; provision managed cloud (AWS RDS, S3, GPUs); stand up Postgres + RDKit cartridge (productionizing the W4 prototype) — Platform Eng
  • Data Engineer #1 onboards; build the molecules / assays / combinations schema from Prateek's W3 draft — Data Eng #1 + Prateek
W6
  • Ingest ChEMBL + PubChem; build ETL pipelines (Airflow / dbt) — Data Eng + Platform Eng
  • 2 Assay Scientists onboard (Prateek's network, HU/IN); confirm collaborator labs; M2-M5 = protocol dev, lab setup & dry runs (productive, not idle; pending the hire-timing decision) — Prateek + Assay Scientists
W7
  • Ingest DrugComb + NCI-ALMANAC; data validation + versioning — Data Eng + Platform Eng
  • Finalize Engine A/B architecture; pick base models (ChemBERTa-2 / MoLFormer / Chemprop); set up W&B tracking — Kourosh + founders
W8
  • Clean, versioned, queryable training tables ready — Data Eng + Platform Eng
  • Encode the first Ayurveda/TCM priors into the data model (formalizing the W4 start) — Prasan + advisors + Nachiket
  • HU/IN contracts live via EOR; bookkeeping live (Pilot / QuickBooks) — Ops Coordinator
TeamW9: Senior ML Engineer + 2nd Data Engineer onboard → core team complete (~11)
AI/MLBegin Engine A build (profiler + combination designer); onboard 2 ML interns (any lab, IP assignment)
Board★ Confirm & appoint the independent director (target; may slip to early Q2)
DataData platform live + queryable for the ML team
Weekly breakdown
W9
  • Senior ML Engineer + 2nd Data Engineer onboard → core team complete (~11) — CEO. If the cold search slips, Kourosh + a strong intern scaffold Engine A so it doesn't stall
  • Kick off Engine A development (profiler + combination designer) — Senior ML + Kourosh
W10
  • Engine A: molecular embedding + property / ADMET prediction — Senior ML
  • ★ Board: confirm & appoint the independent director (target; may slip to early Q2) — CEO
W11
  • Engine A: combination-design logic + Dravya-guṇa priors wired in — Senior ML + Prasan / advisors
  • Onboard 2 ML research interns (any lab, IP assignment) — Senior ML
W12
  • Engine A: first end-to-end run on curated data — Senior ML
  • Quarter review: team in, data platform live, Engine A in progress — all founders
Q2Months 4–6Build the MVP & first lab validation
AI/MLEngine A build continues; rank + score candidate combinations — Senior ML + Kourosh (depends on the M3 senior-ML hire; if it slipped, the gate may land early Q3 — plan slack)
DataPipelines feed clean training data; UMAP representation analysis — Data Eng + Platform Eng
GatePin the in-silico gate pass-criteria now — top-k enrichment / % of known synergies recovered — CEO + Kourosh + Prateek
Wet-labAssay scientists: protocol dev, collaborator-lab setup & dry runs — getting ready for the M6 screens — Prateek + Assay Scientists
Weekly breakdown
W13
  • Engine A: fine-tune on combination data — Senior ML + interns
  • Build the evaluation harness (scaffold / temporal splits, leakage guards, synergy metrics) — Senior ML
  • Pipelines feed clean versioned data; UMAP representation analysis — Data Eng + Platform Eng
W14
  • Engine A produces its first in-silico UTI candidate combinations — Senior ML + Kourosh
  • Prasan + advisors sanity-check candidates for clinical / known-use relevance
  • Define the gate pass-criteria (what 'recovers known synergies' means) — CEO + Kourosh + Prateek
W15
  • Build the in-silico retrospective benchmark (held-out DrugComb / NCI-ALMANAC synergies) — Senior ML + Data Eng
  • Confirm collaborator-lab confidentiality + IP terms before any screening — CEO + Ops + counsel
W16
  • Run the benchmark — does Engine A clear the pre-set bar? — Senior ML
  • Month-end review: can we trust Engine A's rankings? — all founders
Gate★ In-silico gate go/no-go before any lab spend — CEO + Kourosh + Prateek
AI/MLEngine A finalizes the ranked ~30 UTI candidates — Senior ML + Prasan
IPDraft first provisional(s) on the lead combinations — CEO + IP counsel
Weekly breakdown
W17
  • ★ In-silico gate (go/no-go): confirm the benchmark passed; if not, iterate Engine A and slip the lab start — CEO + Kourosh + Prateek
  • Finalize the ranked ~30 UTI candidates — Senior ML + Prasan
W18
  • IP: draft first provisional(s) on lead combinations (AI drafting tool + IP counsel) — CEO + counsel
  • Finalize 384-well checkerboard / FIC protocols — Prateek + Assay Scientists
W19
  • Book collaborator-lab time + order consumables; confirm UTI pathogen panel + MIC standards (CLSI / EUCAST) — Assay Scientists + Ops
  • Engine B scaffolding: pipeline to ingest screen results — Platform Eng + Senior ML
W20
  • Finalize provisional claims around the specific formulations / ratios — CEO + counsel
  • Dry-run the assay workflow + data schema so results return machine-readable — Prateek + Data Eng
IP★ File first provisionals (on the combinations) at the START of the month — before screens / any disclosure (the in-silico design already enables the claims) — CEO + counsel
Wet-labFirst lab validation — 384-well checkerboard / FIC on the lead candidates — Assay Scientists + Prateek
AI/MLEngine B ingests first results → first retrain (the flywheel's first turn) — Senior ML + Kourosh
Weekly breakdown
W21
  • ★ File first provisional patents (on the combinations) now — before screens run / any disclosure — CEO + IP counsel
  • Begin the first 384-well combination screens at collaborator labs (~30 candidates) — Assay Scientists, Prateek lead
W22
  • Run + complete the first 384-well screens — Assay Scientists
  • Keep all screening under confidentiality so it isn't a public disclosure — Ops + counsel
W23
  • Score synergy (FIC; Bliss / Loewe); ingest clean results into the data layer — Assay Scientists + Data Eng
W24
  • Engine B: first feedback retrain on the results — Senior ML + Kourosh
  • Q2 review: MVP nearly there — engines + first validated signal — all founders
Q3Months 7–9MVP validated + the flywheel turns + rolling IP
AI/ML★ MVP validated — both engines working + ranked, lab-validated UTI combinations
Wet-labFlywheel cycle 1: design → test → learn
IPProvisional per newly validated hit
Cadence
Flywheel cadence
  • Typical monthly cycle: AI designs a batch → 384-well screen (collaborator labs) → score synergy (FIC) → ingest results → Engine B retrains → next, better batch.
Wet-labFlywheel cycles; expand the validated set
AI/MLEngine B retrains on accumulating data
IPRolling provisionals (one per promising hit)
Cadence
Flywheel cadence
  • Typical monthly cycle: AI designs a batch → 384-well screen (collaborator labs) → score synergy (FIC) → ingest results → Engine B retrains → next, better batch.
Wet-labMore validated hits (building toward 8–12)
AI/MLImprove ranking from feedback
IPRolling provisionals
Cadence
Flywheel cadence
  • Typical monthly cycle: AI designs a batch → 384-well screen (collaborator labs) → score synergy (FIC) → ingest results → Engine B retrains → next, better batch.
Q4Months 10–12Scale validation, protect IP & expand indications
Decision◆ Go/no-go: if UTI is working, expand beyond UTI to atopic dermatitis + prostate cancer — and hire a cell-culture assay scientist
Wet-labUTI flywheel; accumulate validated hits — Assay Scientists + Prateek
AI/MLModel improvements; Engine A begins designing for atopic dermatitis + prostate cancer — Senior ML + Kourosh
◆ DECISION: expand beyond UTI (AD + prostate cancer) + cell-culture hire ↗
Cadence
Flywheel cadence
  • Typical monthly cycle: AI designs a batch → 384-well screen (collaborator labs) → score synergy (FIC) → ingest results → Engine B retrains → next, better batch.
TeamCell-culture Assay Scientist onboards (mammalian cell assays) — among the assay-dev hires
Wet-labStand up cell-culture assays — keratinocyte / 3-D skin models (atopic dermatitis); prostate-cancer cell lines — Cell-culture Scientist
Wet-labUTI: dose-response + mechanism on lead hits — Assay Scientists
Cadence
Flywheel cadence
  • Typical monthly cycle: AI designs a batch → 384-well screen (collaborator labs) → score synergy (FIC) → ingest results → Engine B retrains → next, better batch.
IP★ Algorithm stays a trade secret — do NOT file a method/algorithm patent (it would publish & reveal the pipeline); revisit only much later, if ever — CEO + counsel
IPBegin FTO on the lead combinations — CEO + counsel
BizMid-point review — incl. results of the multi-indication go/no-go
Q5Months 13–15Deepen validation, multi-indication hits & freedom-to-operate
Wet-labRun cell-culture validation — atopic dermatitis (skin models) + prostate cancer (cell lines) — Cell-culture Scientist
Wet-labUTI flywheel continues; strengthen lead combinations — Assay Scientists
AI/MLFoundation-model refinement across indications — Senior ML
Cadence
Flywheel cadence
  • Typical monthly cycle: AI designs a batch → 384-well screen (collaborator labs) → score synergy (FIC) → ingest results → Engine B retrains → next, better batch.
Wet-labFirst atopic-dermatitis + prostate-cancer hits; validate breadth + replicate key results — Cell-culture + Assay Scientists
IPProvisionals on promising new-indication hits — CEO + counsel
Cadence
Flywheel cadence
  • Typical monthly cycle: AI designs a batch → 384-well screen (collaborator labs) → score synergy (FIC) → ingest results → Engine B retrains → next, better batch.
IPFTO analysis on lead combinations (UTI + new indications); plan conversions — CEO + counsel
Q6Months 16–18Lock the portfolio + non-provisionals + multi-indication portfolio
Wet-labFinalize the 8–12 validated hits
IPFTO complete
IPPrepare non-provisional / conversion drafts
DataData room: assemble validated-hit dossiers
IP★ Begin non-provisional conversions (US)
Wet-lab★ further-indication expansion — building on the Q4-Q5 atopic-dermatitis + prostate-cancer work
BizStart Seed conversations
Q7Months 19–21Non-provisionals + Seed raise
IPConvert lead provisionals → non-provisionals
BizSeed conversations active; data room live
IPNon-provisional / PCT decision window — lock the winners
BizSeed pitch + diligence
BizSeed diligence; refine terms
Q8Months 22–24Close the Seed
BizSeed negotiations
BizDiligence + close mechanics
Biz★ Close ~$4.75M Seed
Goal🎯 Phase 1 done: 8–12 patentable combinations (the IP we file) + the algorithm kept proprietary (trade secret, never patented) + multi-indication validation (UTI, atopic dermatitis, prostate cancer)
📄 DOCSWeek-1 founding & hiring documentsEditable drafts — type to edit · Save draft to keep changes

Full working drafts referenced in Month 1, Week 1. Click into any document to edit it directly; Save draft stores your edits in this browser, Revert restores the original. Finalize with counsel before signing/posting.

Founders' AgreementFull contract draft — equity, per-founder roles, vesting, board, IP & leaver terms

DRAFT v1 for founder review — not yet executed. To be finalized into binding documents by corporate counsel (Role 1b). Bracketed items [ … ] are decisions for the founders to agree. Not legal advice.

Founders' Agreement

Company: Biovizions, Inc., a Delaware C-corporation (the "Company").
Effective date: [ ____________ , 2026 ]
Founders: Nachiket Shankar, Prateek Shetty, Kourosh Zarringhalam, and Prasan Shankar (each a "Founder").

This Founders' Agreement records the founders' mutual understanding on equity, roles, vesting, decision-making, IP, and departure terms. Where it touches matters governed by the Company's Certificate of Incorporation, Bylaws, stock-purchase agreements, or stock plan, those formal documents control once executed.

1 · Formation & ownership

The Company is already incorporated in Delaware (Nachiket Shankar, sole stockholder to date). Founder stock will be issued to the remaining three founders under Restricted Stock Purchase Agreements (RSPAs) with the vesting in Section 4. A 10% option pool (ESOP) is reserved for employees, advisors, and future hires.

HolderRoleEquity %VestingNotes
Nachiket ShankarCEO[ __ ]%4yr / 1yr cliffsole founder to date
Prateek ShettyCo-founder — Assay/Lab + Data[ __ ]%4yr / 1yr cliffshares to be issued
Kourosh ZarringhalamCo-founder — AI/ML[ __ ]%4yr / 1yr cliffshares to be issued
Prasan ShankarCo-founder — Clinical[ __ ]%4yr / 1yr cliffshares to be issued
ESOP (option pool)10%reserved for hires/advisors
TOTAL100%

Guidance: equal-ish splits among full-time co-founders are common, adjusted for role, time commitment, idea/IP origination, and capital contributed. Founders to agree and complete the table.

2 · Roles & responsibilities

Each founder's mandate, responsibilities, and decision authority:

Nachiket Shankar — Co-founder & CEO

Mandate: Company vision, strategy, and execution; the face to investors.

  • Lead the pre-seed and Seed fundraising and investor relations.
  • Set company strategy and the 24-month roadmap.
  • Own the cap table, the board, and company finances / budget.
  • Recruit and manage the team; set culture.
  • Drive computational-biology direction.
  • Own the legal / IP relationships, with outside counsel.

Decision authority: final call on strategy, fundraising, budget, and senior hires (with founder input); signs contracts on the Company's behalf.

Prateek Shetty — Co-founder & Head of Assay Development / Lab + Data

Mandate: the wet-lab science and the data foundation.

  • Design and oversee the combinatorial / synergy assays (checkerboard, FIC).
  • Manage collaborator-lab relationships and lab operations (acts as lab manager).
  • Lead the assay-development scientists.
  • Own data quality and the wet-lab → ML training-data pipeline.
  • Co-own the cheminformatics data modeling (with Nachiket).

Decision authority: assay design, lab partnerships, and the wet-lab budget.

Kourosh Zarringhalam — Co-founder & Head of AI/ML (Chief AI Scientist)

Mandate: scientific direction of the AI engines.

  • Set the ML / algorithm strategy (the two engines, the priors approach).
  • Provide scientific oversight of the senior ML engineer and the interns.
  • Own model architecture and validation methodology.
  • Bridge academia ↔ company. Part-time (professor).

Decision authority: AI/ML scientific direction and model approach.

Prasan Shankar — Co-founder & Chief Medical / Clinical Officer

Mandate: clinical grounding and the traditional-medicine knowledge base.

  • Clinical / translational strategy.
  • Ayurveda / Dravya-guṇa domain knowledge and priors (with the advisors).
  • Indication selection and clinical relevance.
  • Clinical insight and future-trial access via the integrative-medicine hospital.
  • Regulatory / clinical advisory.

Decision authority: clinical direction, indication priorities, and the traditional-medicine knowledge base.

3 · Time commitment

  • Nachiket Shankar (CEO) — full-time.
  • Prateek Shetty — full-time.
  • Kourosh Zarringhalam — part-time (university professor); ~[ __ ] hrs/week of scientific direction.
  • Prasan Shankar — part-time (clinical); ~[ __ ] hrs/week.

Founder cash compensation, if any, is set separately by the Board and reflected in the budget; it is not part of this Agreement.

4 · Vesting

  • Standard 4-year vesting with a 1-year cliff; monthly vesting thereafter.
  • Unvested shares are subject to repurchase by the Company at the lower of cost or fair value on departure.
  • Each Founder files an 83(b) election within 30 days of share issuance (time-critical tax step).
  • [ Optional: vesting credit for pre-incorporation work — to agree. ]

5 · Decision rights

Day-to-day decisions sit with the relevant function lead (Section 2). The following require all-founder consultation and, where noted, Board approval: fundraising & cap-table changes, the annual budget and major spend, equity grants, senior/key hires, major partnerships and contracts, and any change to this Agreement. (Full matrix maintained in the Founders' Governance doc.)

6 · Board of Directors

  • Initial board kept small and founder-controlled.
  • At least one independent director to be identified and appointed within the first 3 months (target: a biotech/pharma operator or drug-discovery scientist).
  • Target structure: CEO + one co-founder + one independent; the lead investor typically joins at Seed.

7 · Intellectual property

Each Founder signs the Company's Confidential Information & Invention Assignment Agreement (CIIA). All inventions, code, models, data, and know-how created for the Company are assigned to the Company. No Founder retains any personal claim to Company IP. On departure, all IP remains with the Company.

8 · Leaver provisions

These terms live in this Agreement and the Right-of-First-Refusal & Co-Sale Agreement (reflected in the bylaws). Plain-language summary:

TermWhat it meansWhy it matters
Good vs bad leaverA good leaver departs for acceptable reasons (death, disability, termination without cause, mutual agreement) and keeps vested shares. A bad leaver departs for cause (misconduct / breach) or resigns early — forfeits unvested shares, and sometimes vested ones are repurchased at cost.Stops someone who walks early or is fired for cause from keeping equity they didn't earn.
Vesting accelerationUnvested shares vest early on a trigger. Single-trigger = on a company sale; double-trigger (the market norm) = a sale AND being let go without cause shortly after.Sets what founders/staff receive if you're acquired; affects retention and acquirer terms.
Right of first refusal (ROFR)If a holder wants to sell shares to an outsider, the Company + other shareholders may buy them first on the same terms.Keeps the cap table clean — you control who becomes an owner.
Co-sale (tag-along)If a founder sells to an outsider, the other shareholders can join and sell their proportional share on the same terms.Protects the other founders — no one is left behind when someone cashes out.
Drag-alongIf a defined majority approves a company sale, they can require the minority to sell on the same terms.Lets a clean acquisition close — a small holdout can't block a deal the majority wants.
IP on departureAll IP created stays assigned to the Company (per the CIIA); a departing founder has no claim.The company's core asset — the IP — can't leave with a founder.

9 · Confidentiality

Each Founder keeps Company confidential information secret during and after involvement with the Company, per the CIIA.

10 · General

  • Governing law: Delaware.
  • This Agreement may be amended only in writing signed by all Founders (subject to Board/stockholder approval where required).
  • May be executed in counterparts.
  • If any provision conflicts with the Company's formal organizational documents, those documents control.

Signatures

Nachiket Shankar  ____________________________   Date ____________

Prateek Shetty  ____________________________   Date ____________

Kourosh Zarringhalam  ____________________________   Date ____________

Prasan Shankar  ____________________________   Date ____________

CIIA / IP-Assignment AgreementConfidential Information & Invention Assignment — founders + all hires

DRAFT — standard Confidential Information & Invention Assignment Agreement for founders and all hires. To be finalized by corporate counsel (Role 1b) and executed via Clerky / DocuSign. Not legal advice.

Confidential Information & Invention Assignment Agreement (CIIA)

Between Biovizions, Inc. ("Company") and [ Full name ] ("I" / "me"), effective [ date ], in consideration of my employment, founder role, engagement, or continued service and the equity / compensation I receive.

1 · Confidential information

I will hold in strict confidence and not disclose or use, except for the Company's benefit, any Confidential Information — including the Company's models, algorithms, code, datasets, chemical / biological data, combination designs, screening results, business and financial information, and third-party information the Company must keep confidential — during and after my relationship with the Company.

2 · Assignment of inventions

I assign to the Company all right, title, and interest in all inventions, discoveries, developments, software, models, data, and works of authorship that I make or conceive, alone or with others, during my relationship with the Company that (a) relate to the Company's business or research, or (b) are made using Company time, resources, or Confidential Information ("Company Inventions"). I will promptly disclose all such Inventions to the Company.

3 · Prior inventions

Inventions I made before this relationship and wish to exclude are listed in Exhibit A. If I list none, I represent there are none. If I incorporate any prior invention into a Company product or Invention, I grant the Company a non-exclusive, royalty-free, perpetual license to use it.

Exhibit A — Prior inventions (list, or write "None"): [ ____________________________________ ]

4 · Excluded inventions (statutory carve-out)

This assignment does not apply to an invention that I developed entirely on my own time without using Company equipment, supplies, facilities, or Confidential Information, and that does not relate to the Company's business or actual / anticipated research, and does not result from work I performed for the Company — consistent with applicable state law.

5 · Further assurances

I will help the Company secure and enforce its rights in Company Inventions (e.g., signing patent and copyright documents). I irrevocably appoint the Company as my attorney-in-fact to execute such documents if I am unavailable.

6 · Return of materials

On termination of my relationship, I will return all Company property and materials (including all copies and data) and will not retain any Confidential Information.

7 · No conflicting obligations

My performance does not and will not breach any agreement with a prior employer, university, or third party. I will not improperly use or disclose any third party's confidential information or trade secrets, nor bring any onto Company premises or systems.

Note for academic founders / interns: confirm there is no conflicting university IP obligation; keep interns arms-length from any founder's own lab to keep IP clean.

8 · General

  • Governing law: [ Delaware / state of employment ].
  • This Agreement survives the end of my relationship with the Company.
  • If any provision is unenforceable, the rest remains in effect.

Signature

Signature ____________________________   Print name ____________________________   Date ____________

For Biovizions, Inc.: ____________________________   Nachiket Shankar, CEO   Date ____________

Senior AI/ML Engineer — Job DescriptionFounding technical hire · post in Week 1

DRAFT job post — edit freely before posting. Comp benchmarks are US, on the 24-month plan.

Senior AI / Machine Learning Engineer (Founding)

Type: Full-time, founding technical hire · Comp: ~$11K/mo ($10–12K band) + 1% equity (4-yr vest / 1-yr cliff) · Location: remote-friendly (US-benchmarked) · Reports to: CEO; works closely with Prof. Kourosh Zarringhalam (AI/ML co-founder).

About Biovizions

We design safe, low-dose drug + phytochemical combinations by encoding Ayurveda / TCM priors into modern molecular ML. You'd be employee #1 on the technical side, turning that thesis into working models.

What you'll build

  • Engine A — profiler + combination designer: embed drugs and phytochemicals in a shared learnable chemical space; predict pharmacological / ADMET / PK profiles; design combinations under dosage-boosting and toxicity-balancing rules.
  • Engine B — active-learning loop: ingest high-throughput screen results, propose the next experiments (active learning / Bayesian optimization), and reinforce a foundation model for combinatorial discovery.
  • Own the molecular ML stack end-to-end — data, training, evaluation, the chemical-space map — with MLOps / platform deploying alongside you.

You have

  • Hands-on experience training molecular models — chemical language models (ChemBERTa-2, MoLFormer) and/or graph neural nets (Chemprop / D-MPNN) — and doing property / ADMET prediction.
  • A PhD in ML-for-drug-discovery / computational chemistry / cheminformatics, or equivalent TechBio experience (e.g., Recursion, Atomwise, Insilico, Iambic, Genesis, Valence).
  • Strong Python, PyTorch, RDKit, and experiment tracking (Weights & Biases).
  • First-technical-hire DNA: scrappy, owns the whole problem, thrives with founders and an academic lab.

Nice to have

  • Active learning / experimental design; multi-target or natural-product modeling; attention-MIL on screening data.

What we offer

  • Founding equity and real ownership of the core IP; a novel scientific problem; a tight, senior founding team.

How to apply

Email [ careers@biovizions.com ] with your CV / GitHub / Google Scholar and a few lines on a molecular model you've trained. [ Edit before posting. ]

Operations & People Coordinator — Job DescriptionFractional ops hire · post in Week 1

DRAFT job post — edit freely before posting.

Operations & People Coordinator (Cross-Border) — fractional / contract

Type: Contract, ~20–25 hrs/week · Comp: ~$3.5K/mo (cash) · Reports to: CEO · Path: to Head of Ops at Seed.

About the role

Be the operational backbone of an early biotech-AI startup with a cross-border team (US, Hungary, India). You'll keep hiring, payroll, compliance, and the data room running so the founders can focus on the science.

What you'll do

  • Run cross-border onboarding: issue / collect signed contractor & employee agreements (incl. IP assignment, drafted by counsel); set each person up in the right structure (US, Hungary services entity, India contractors).
  • Manage payroll & payments via Deel / Rippling; process Hungary-entity invoices; everyone paid on time and compliantly.
  • Bookkeeping liaison with the accountant (Pilot / QuickBooks); keep compliance, IP-assignment tracking, and the investor data room current.
  • Run grant logistics (Google / Anthropic / Nvidia credits and any non-dilutive grants).
  • Own the ops stack: Deel, Clerky, Carta, Pilot / QuickBooks, DocSend.

You have

  • 2+ years in startup ops / people / EA roles; comfort with cross-border contractors and payroll tools.
  • Highly organized, discreet with confidential information, and self-directed.
  • Bonus: experience with Deel / Rippling, Carta, and early-stage fundraising logistics.

How to apply

Email [ careers@biovizions.com ] with your CV and a short note on a cross-border ops process you've run. [ Edit before posting. ]

Data Room — Index & ChecklistFolder structure + what goes in each · stand up in Week 1

The data-room shell to stand up in Week 1 (DocSend or a structured Google Drive). Create these folders now; fill as documents are produced. "Status" tracks readiness.

Data room — index & document checklist

FolderWhat goes hereStatus
01 · Corporate & formationCertificate of Incorporation (DE C-corp), Bylaws, board & stockholder consents, EIN, good-standingCert ✓ · rest to do
02 · Cap table & equityCarta cap-table export, founder RSPAs, 83(b) confirmations, Stock Plan (10% ESOP), 409A, SAFE notesTo do (W2)
03 · Intellectual propertyCIIAs (all founders + hires), provisional patents, FTO memos, invention disclosures, trademark filingCIIA + TM in W1
04 · Founders & teamFounders' Agreement, governance doc, org chart, JDs, offer letters, advisor agreements (incl. IP)Drafts in W1
05 · FinancialsBudget model, burn / runway, bank statements, accounting (Pilot / QuickBooks), grant applicationsModel ✓ · rest to do
06 · Science & dataData sources & licenses (ChEMBL / PubChem / DrugComb / NCI-ALMANAC), model-architecture overview, validation plan, assay protocolsBuilding W3–W4
07 · Legal & complianceMutual NDA, collaborator-lab agreements (with IP terms), EOR / contractor agreements, insurance (GL, D&O), Claude API ZDR / BAATo do
08 · Pitch & strategyPitch deck, 24-month roadmap, market & competitive notesDeck ✓ · roadmap ✓

Keep a clean, read-only investor view; stage sensitive IP (raw data, model weights, unfiled invention details) in a restricted sub-folder shared only under NDA.