2025 State of TechBio

Hi everyone,

The 2025 State of TechBio survey results are now live, and I wanted to share what we learned from 200+ members of the Bits in Bio community.

See the results

Who we surveyed

Our respondents were overwhelmingly industry-focused (82% vs. 18% academia), computational (top roles: Comp Bio 35%, Data Science 24%, Software Engineering 21%), and split between early-stage startups and large organizations. The typical respondent is highly educated (45% PhD, 31% Master’s), US-based (79%, mostly CA and MA), and hands-on technical (69% write code regularly).

What Changed This Year

Looking at year-over-year trends, three shifts stand out:

AI tool usage consolidated dramatically. Claude saw a 28-point jump (32%→60%), while proprietary/in-house models dropped 16 points (60%→44%). OpenAI usage also grew 10 points—from 52% using GPT-4/GPT-4o to 62% using GPT-5, o1, and newer models.

The field got more hands-on technical. Those who say they do ML work every day increased 13 points (15%→28%), while those who “never” do ML dropped 11 points (25%→14%). Daily development of developer-facing tools and libraries jumped 12 points (16%→28%). Biotech continues to incorporate more and more software.

Market sentiment cooled significantly. The “very bullish” camp shrank from 20% to 9% (−11 points), while “somewhat bearish” grew from 15% to 27% (+12 points). We’re still seeing more bulls than bears overall, but people are still feeling the extended biotech slump.

What People Are Actually Using

The survey also mapped the current TechBio stack in detail:

• Languages: Python dominates (93%), followed by Bash (57%), R (42%), and SQL (40%)

• ML Frameworks: Scikit-Learn (67%) and PyTorch (64%) lead, with OpenAI (41%) and HuggingFace (38%)

• Cloud: AWS is the overwhelming choice (67%), with GCP a distant second (34%)

• Data formats: CSV/TSV (71% input, 57% output), FASTA (50%), JSON (49%)

• ML applications: Writing code (65%), literature search (40%), target ID (32%)

Community Insights

The Bits in Bio community continues to grow primarily through word of mouth (68%), with Slack as the most valued offering (66%). However, engagement is mostly “drop-in”—37% visit weekly or a few times per month, vs. just 11% daily.

Explore the Full Dataset

The complete results include interactive charts on everything from salary ranges to lab automation vendors to LIMS preferences. You can also download the raw data (CSV) and a full PDF report.

👉 Explore the 2025 State of TechBio Survey

We’re grateful to everyone who took the time to share their experiences, and to Gunderson Dettmer, Ropes & Gray, and AWS for sponsoring this work. If you have feedback or suggestions for future surveys, please reach out—we’d love to hear from you.

P.S. If you found these results useful, consider sharing them with colleagues who might benefit. The more participation we get next year, the richer the insights become.