2025 State
of TechBio
The 2025 State of TechBio Survey aims to provide data to those who want to improve the software and tools used in the field. We hope that the community will discover insights on attitudes, tools, and environments influencing science and software today.

Feedback
We welcome feedback! Please reach out to Shaq Vayda with your suggestions. For any technical issues or feedback concerning this website, please reach out to survey@nitro.bio.

YoY Trends
2024 → 2025Model Usage Shifts
"Build in-house" AI has cooled off, while Claude surged into the top tier of day-to-day model usage.
More Hands-On Technical
The field is becoming more hands-on technical—daily ML and daily tool-building both jumped.
Sentiment Turned Cautious
The "very bullish" camp shrank while bearishness rose meaningfully.
Survey Results are
grouped into 4 sections:


BiB Community
BiB is a word-of-mouth, Slack-first network: people join through people, and most value comes from the always-on community, room to continue to increase participation
- •68% of you discovered BiB via word of mouth (vs 11% meetups, 8% LinkedIn)
- •66% say Slack is BiB's most useful offering (vs 47% in-person meetups, 37% job board)
- •Though engagement today is mostly "drop-in": 37% visit weekly/a few times per month (vs 11% daily)
How did you hear about Bits in Bio?
194 RespondentsHow frequently would you say you go on the Bits in Bio Slack Channel during a typical month?
200 RespondentsWhich part of Bits in Bio do you find most useful?
202 RespondentsDo you consider yourself a member of the Bits in Bio community?
202 Respondents
Professional Experience
TechBio talent is a blended market: overwhelmingly industry-based, strongly computational, and split between early-stage startups and bigger-orgs
- •82% are in (or seeking) industry roles (vs 18% academia)
- •Top roles skew compute: 35% Comp Bio/Bioinformatics, 24% Data Scientist, 21% Software Engineer, 20% ML Eng/Researcher
- •Org size is barbelled: 35% at 2–19 employees, while 17% are at 5,000+
Do you consider yourself to be working (or seeking work) in industry or academia?
194 RespondentsWhich of the following best describes your current employment status?
206 RespondentsWhat sources do you use to keep up to date in your field?
206 RespondentsWhich of the following best describes your current role?
206 RespondentsWhat industry does your organization operate in?
206 RespondentsApproximately how many people are employed by the company or organization you currently work for?
206 RespondentsWhich of the following best describes your current role/years of experience?
175 RespondentsWhat annual salary range do you fall into?
197 Respondents
Technical Experience
LLMs have become the default work surface in TechBio: most respondents touch ML, and the most common "ML win" is accelerating human workflows (coding + knowledge search) as much as modeling biology.
- •67% say their work has involved ML (+ 23% curious but not yet)
- •Most common ML objective: 65% use ML for writing code (vs 40% literature search; 32% target identification/selection)
- •Tooling is LLM-led: OpenAI 62% and Claude 60% among those regularly working with ML models
Which persona do you most identify with?
202 RespondentsHas your work involved machine learning?
206 RespondentsWhich of the following machine learning models do you regularly work with?
129 RespondentsWhich of the following design objectives do you regularly work on using machine learning?
136 RespondentsHas your work involved lab automation?
206 RespondentsWhich lab equipment vendors do you regularly work with?
54 RespondentsWhich of the following design objectives do you regularly work on using lab automation?
58 RespondentsDo you code?
206 RespondentsWhich of the following languages do you use day-to-day?
162 RespondentsHow often do you personally do data analysis for work?
166 RespondentsHow often do you personally write or work on data engineering/bioinformatics pipelines for work?
166 RespondentsWhich of the following frameworks do you use (if any) for your data engineering tasks?
153 RespondentsHow often do you personally develop on developer-facing tools/libraries for work?
165 RespondentsHow often do you personally develop on scientist-facing tools or interfaces?
166 RespondentsHow often do you personally do machine learning for work?
166 RespondentsWhich of the following machine learning tools (frameworks, libraries, models, etc) do you work with?
134 RespondentsHow often do you personally program web interfaces for work?
166 RespondentsWhich databases have you worked with extensively?
153 RespondentsWhich cloud platforms have you extensively developed with?
166 RespondentsWhich of the following data sources do you use extensively?
206 RespondentsWhen you are using data from another source, what formats does it come in (input data)?
206 RespondentsWhen you are storing data for analysis or later use which formats do you use (output data)?
206 RespondentsWhat tools and libraries do you use to visualise data?
206 RespondentsWhich of the following lab information management systems and electronic lab notebooks do you use?
206 Respondents
Respondent Demographics
BiB is a highly educated, coastal-US respondent base that's "cautiously mixed" on market outlook—but shows strong advocacy via high recommendation scores.
- •Geography is concentrated: 79% US; within US, 49% CA and 22% MA
- •Education is very advanced: 45% PhD/doctoral, 31% master's
- •Sentiment is mixed-tilt-bullish: 33% somewhat bullish, 27% somewhat bearish, 25% neutral
