Anmol Gorakshakar

New York, NY · agorakshakar@gmail.com · LinkedIn · Google Scholar

Summary

Platform engineer and computational biologist with 8+ years closing the gap between scientific research and production systems. I build the infrastructure that puts research into the hands of customers — scoring engines, internal developer platforms, and data pipelines that eliminate toil and prevent errors at scale. At Viome, I own the system that delivers every microbiome health score to customers and clinicians. At GeneCentrix, I achieved a 1,000× speed improvement on a core algorithm and got it integrated into a commercial product. My edge is knowing both the biology and the engineering well enough to build things that actually ship.

Core Skills

Languages: Python (advanced), SQL, Bash, R
Frameworks & APIs: Flask · NumPy · Pandas · Scikit-learn · PyTorch
Platform & Infrastructure: Kubernetes (AKS) · Terraform · KEDA · CircleCI CI/CD · Apache Airflow · Azure · AWS · PostgreSQL
ML / Scoring: Score inference pipelines · RNA-seq normalization · Molecular docking scoring
Data Engineering: Relational schema design · ETL pipelines · Composable pipeline architecture · Audit-log systems · Deterministic UUID / reproducible scoring
Bioinformatics: Systems biology · Multiomics (RNA-seq, microarray, exome seq) · Protein structure modelling · Molecular docking · Microbiome scoring
Domain Expertise: Microbiome / gut health · Drug discovery · Precision medicine · Clinical data infrastructure

Professional Experience

Platform Engineer, Data Analytics & Solutions

Viome Life Sciences · New York, NY · March 2025 – Present · Full-time

Viome's scoring infrastructure was fragile: models deployed manually to VMs, a 3-day data sync blocking downstream work, and silent errors in customer scores that had gone undetected for years. I was brought in to fix that — architecting and owning the Viome Scoring Engine (VSE) end-to-end, the central platform now delivering every microbiome health score to Viome's customers and clinicians.

Cut the ODM Parquet sync from 3 days to 8 hours by rewriting O(n²) parsing to O(n), unblocking data engineers from a weekly bottleneck that delayed every downstream pipeline.
Replaced all legacy VMs with always-on Kubernetes infrastructure for scoring (AKS, Terraform, KEDA, CircleCI CI/CD) and a 5-step model deployment pipeline (build → package → promote → validate → rollout), giving the team a reliable, auditable path from model to production with deterministic scoring across dev, staging, and prod.
Built DASH, Viome's internal scoring release platform, to standardize the full release workflow behind a single make install — eliminating a fragmented, error-prone multi-step process. Ships compiled CLI tools (scoring, das-herald, translate-ids) used daily across the DAS team.
Built metadata-labeler to replace ad-hoc metadata scripts with a versioned, append-only labeling pipeline, giving DS, QA, and Science a live metadata view instead of manual SQL queries and enabling reproducible metadata across all scoring runs.
Turned a 30+ minute manual SQL audit into a self-service tool by productionizing Score Detail Tool from a single-user Streamlit prototype into a multi-user Flask/HTMX app serving 20–50 people across DS, QA, Support, Science, and Product.

Scientist, Systems Biology

Viome Life Sciences · New York, NY · May 2022 – March 2025 (2 years 10 months) · Full-time

Viome had no standardized tooling for developing or validating microbiome scores — scientists worked from ad-hoc scripts with no shared evaluation framework and no way to explain scores to the rest of the business. I built the foundational infrastructure that became the core of VSE and the Score Detail Tool.

Built the algorithmic foundation of VSE: authored the 2 internal Python packages covering score computation, normalization (CLR, CPM, quantile), and model evaluation that became VSE's core production logic.
Made microbiome scores explainable for the first time by building Score Detail Tool v1 (Streamlit) — collapsing a 30+ minute multi-database SQL audit into a self-service lookup adopted across DS, QA, and Science.
Consolidated fragmented metadata scripts into a single repeatable pipeline (MetadataGenerator), unifying kit identifier resolution, questionnaire joins, and binarization, enabling consistent metadata across the entire scoring workflow.
Built a POC for Peacock (Viome's AI chatbot) — designed the context layer surfacing biological data to the LLM backend, enabling the first AI-assisted health score explanations for customers.
Maintained Correlational Controls drift monitoring, giving the executive team the data to make prioritization decisions on score design.
Documented all SB tools as a Read the Docs site, enabling independent score development across DS and SB teams without relying on tribal knowledge.

Research Scientist

GeneCentrix, Inc. · New York City Metro · May 2020 – April 2022 (2 years) · Full-time

GeneCentrix needed to productize its computational biology IP — the core algorithm was too slow for real-world use and had no delivery layer for external partners or customers.

Achieved a >1,000× speed improvement on the HR-Algorithm by rebuilding it as a standalone Python package and relational database from scratch — making it fast enough to use commercially. The platform was integrated into Molsoft ICM v3.9-3 and documented on their New ICM Features page, putting it in the hands of pharmaceutical teams worldwide.
Advanced 5+ small molecule leads to preclinical CRO validation by leading a two-part AbbVie collaboration targeting neuropathic pain — building a high-throughput in-silico screening platform and an engineered nanobody-toxin delivery system.
Built a Flask REST API and Molsoft ICM integration layer on top of the algorithm, enabling external partner access without requiring direct software installs.

Earlier Research

Research Intern

Homi Bhabha Centre for Science Education (HBCSE), TIFR · Mumbai, India · December 2016 – February 2018

Investigated mTOR signalling in Chlamydomonas reinhardtii; work contributed to a peer-reviewed publication in FEBS Letters (33 citations).

Education

New York University · MS, Biotechnology · 2018 – 2020 · GPA 3.9 / 4.0

Vellore Institute of Technology · MSc, Biotechnology · 2015 – 2017 · Grade 8.14 / 10.00

St. Xavier's College, Mumbai · BSc, Microbiology · 2012 – 2015

Publications

50 total citations · h-index 3

Patridge E, Gorakshakar A, Molusky MM, Ogundijo O, Julian C, Hu L, et al. Altered gut microbial functional pathways in people with irritable bowel syndrome enable precision health insights. BMC Gastroenterol. 2025;25(1):823. (5 citations)
Patridge E, Gorakshakar A, Molusky MM, Ogundijo O, Janevski A, et al. Microbial functional pathways based on metatranscriptomic profiling enable effective saliva-based health assessments for precision wellness. Comput Struct Biotechnol J. 2024;23:834-842. (7 citations)
Patridge E, Gorakshakar A, Molusky MM, Ogundijo O, Julian C, Hu L, et al. Sub-optimal activity of gut microbiome functional pathways increases the odds of irritable bowel syndrome in a large adult human population. 2024.
Julian C, Shen N, Molusky M, Hu L, Gopu V, Gorakshakar A, Patridge E, et al. The effectiveness of precision supplements on depression symptoms in a US population. medRxiv. Published April 18, 2023. (2 citations)
Upadhyaya S, Agrawal S, Gorakshakar A, Rao BJ. TOR kinase activity in Chlamydomonas reinhardtii is modulated by cellular metabolic states. FEBS Lett. 2020;594(19):3122-3141. (33 citations)
Upadhyaya S, Agrawal S, Gorakshakar A, Rao BJ. Modulation of TOR kinase activity in Chlamydomonas reinhardtii: effect of N-starvation and changing carbon pool. bioRxiv. Published 2018. (3 citations)