Proof of concept for biotech startups: what to run in-house and what to outsource
Biotech proof of concept for Indian startups: which experiments to keep in-house, which to outsource, what BIRAC BIG funds and a worked budget.
A biotech proof of concept shows that the critical function works in the lab. Startups should keep the core experiment, which holds the know-how, in-house and outsource routine analytics, specialised instruments and regulated studies. BIRAC's BIG gives up to ₹50 lakh over 18 months, and outsourcing is one of its budget heads.
A biotech startup with a promising idea and eighteen months of grant money faces the same question every week: do we run this experiment ourselves, or pay someone to run it? Buying equipment and hiring for every step burns money and time. Outsourcing everything leaves the startup with reports but no know-how.
The answer is rarely all one or the other. This guide sets out what proof of concept means in biotech, a decision table for each type of work, what India's main proof-of-concept grant allows, how to protect your IP when you outsource, and a worked budget.
What does proof of concept mean in biotech?
Proof of concept means experimental evidence that the critical function works as predicted. For a diagnostic, it is detecting the target at the sensitivity you need. For an enzyme, it is reaching target activity under realistic conditions. For a production strain, it is making the compound at a measurable titre.
In technology readiness terms, this is TRL 3: experimental proof of concept, before components are put together and validated as a system at TRL 4. BIRAC's own readiness scale puts proof of concept in the same place; TRL for biotech products explains BIRAC's sector scales, and moving from TRL 3 to TRL 4 covers the step after.
Which work should stay in-house and which can go out?
The deciding questions are simple. Does this work create know-how that makes us valuable? Do we need to repeat it often? Can we get the equipment and skill cheaply ourselves?
| Type of work | Keep in-house when | Outsource when | Typical choice |
|---|---|---|---|
| Core method development and optimisation | Almost always; this is your know-how | Only for a specific technique you will never repeat | In-house |
| Routine analytics (HPLC, basic assays) | You run it weekly and the instrument is available at your incubator | You need it a few times, or need validated methods | Often outsourced early |
| Sequencing and omics | Rarely, unless sequencing is your product | Most cases; providers have scale | Outsourced |
| Specialised characterisation (mass spectrometry, microscopy, structural studies) | Rarely | Most cases | Outsourced, or shared facility |
| Strain or cell line banking | You need fast access to working stocks | For master banks and long-term storage | Mixed |
| Animal studies | Almost never at startup stage | When needed, with an approved facility | Outsourced |
| Studies needing GLP or other certification | Almost never at startup stage | When the data will go to a regulator | Outsourced |
| Pilot-scale production | Rarely | When you need material beyond bench scale | Outsourced |
Shared instruments at incubators and institutional core facilities sit between the two columns. AIIMS New Delhi, for example, describes its Centre for Medical Innovation and Entrepreneurship as a bio-incubator under BIRAC's BioNEST scheme, alongside a centralised core research facility. Using a shared instrument with your own hands keeps the know-how and avoids buying the machine.
Keep it or send it out
Own the critical experiment; buy the services around it.
What does BIRAC BIG allow?
BIRAC's Biotechnology Ignition Grant gives up to ₹50 lakh over up to 18 months to take an idea to proof of concept, released in milestone-based instalments. The current BIG guidelines (version 9, October 2025) list the budget heads, including manpower, consumables, travel, outsourcing, contingency and equipment, without percentage caps; budget caps are checked during the site visit. The earlier version 8 guidelines (July 2020) did set limits, and they remain a useful guide to what reviewers consider balanced:
| Budget head | Limit in the version 8 guidelines (2020) |
|---|---|
| Outsourcing | Not more than 30% of project cost |
| Equipment | Not more than 30% of project cost |
| Manpower | About 30% of project cost, up to ₹50,000 a month per person |
| Travel | Up to ₹1.5 lakh |
| IP costs (drafting, filing) | Up to ₹1.5 lakh |
| Contingency | Not more than 5% of project cost |
Version 8 allowed outsourcing for project services not available at the incubator or the startup's own lab, and short-term specialised consultancy. It did not fund systematic clinical trials, field trials or late-stage validation and certification, and it required individual applicants to be incubated and companies to have a functional in-house lab or incubation. Read the current call document before you build the budget. BIRAC BIG grant and TRL covers the evidence reviewers look for.
Even without a published cap, keeping outsourcing to around 30% is a useful discipline. It pushes the core work in-house, which is where it belongs anyway.
Which approvals apply before you start?
Some experiments need approvals whether you run them yourself or outsource them. Work with genetically engineered organisms is regulated under the Manufacture, Use, Import, Export and Storage of Hazardous Micro-organisms, Genetically Engineered Organisms or Cells Rules, 1989, made under the Environment (Protection) Act, 1986, and the Department of Biotechnology's Regulations and Guidelines for Recombinant DNA Research and Biocontainment, 2017, with review by an Institutional Biosafety Committee; DBT's biosafety portal lists the rules and the IBSC handbook. Animal work needs an approved facility and ethics committee. Work on human samples needs ethics approval.
Outsourcing does not move the responsibility for these approvals off your project. Ask each provider which approvals cover the work and get copies before you send material.
How do you protect IP when you outsource?
Every outsourced experiment sends something out of your lab: a strain, a sequence, a protocol or data. Before that happens:
- Sign a confidentiality agreement and a service agreement, or a material transfer agreement for biological material.
- State in writing that all results, data and inventions arising from the work belong to you.
- Send only what the provider needs. A provider measuring titre does not need your strain construction method.
- Ask for raw data, not only a summary report, and the method used.
- File a patent application before results are shared widely, if the result is patentable.
The know-how question matters for later deals. If a licensee or investor asks how your process works and the answer is "the contract lab knows", you have weakened your own position. Know-how licences in India explains why the unwritten detail is often worth more than the patent.
Before any material leaves your lab
A worked budget
Illustration with invented figures. A Bengaluru startup is proving that an engineered enzyme breaks down a food-industry waste stream, under a BIG grant, with a project cost of ₹48.4 lakh.
| Item | In-house or outsourced | Cost |
|---|---|---|
| Enzyme engineering and activity optimisation | In-house, incubator lab | ₹12,00,000 consumables |
| Scientist and research associate, 18 months | In-house | ₹14,40,000 |
| Bench equipment (shaker, small fermenter) | In-house | ₹8,00,000 |
| Sequencing of variants | Outsourced | ₹3,00,000 |
| Product analysis by LC-MS | Outsourced | ₹5,50,000 |
| Pilot hydrolysis trial at 20 litres | Outsourced | ₹4,50,000 |
| Patent filing | Outsourced | ₹1,00,000 |
| Total | ₹48,40,000 |
Outsourced services, excluding the patent filing, come to ₹13 lakh, about 27% of the project, inside the 30% that the 2020 guidelines allowed. Equipment is about 17%, and manpower about 30%, at ₹40,000 a month for each of the two people. The core engineering stays in the startup's hands. How each line is classified and capped depends on the guideline in force, so confirm with BIRAC.
Common mistakes
- Outsourcing the core experiment because it is faster, then not knowing how to repeat it.
- Buying an instrument used twice a month when a shared facility has one.
- Sending a strain to a provider without a written IP and confidentiality agreement.
- Accepting a summary report without raw data.
- Planning a BIG budget around late-stage validation, which BIG is not designed to fund.
My view: own the experiment that would be hardest to explain to someone else, and outsource the rest. I cannot give a standard split for every field, because a diagnostic, a strain and a biologic need very different services, so treat the table above as a starting point.
When you need lab capacity beyond your own bench, Zernyx Labs does prototyping and lab R&D for biotech, pharma and cosmetics.
Common questions
What counts as proof of concept in biotech?
Experimental evidence that the critical function of the technology works as predicted, such as an enzyme reaching target activity, a strain producing the compound or an assay detecting the analyte at the needed sensitivity. In TRL terms this is TRL 3, experimental proof of concept.
How much of a BIRAC BIG project can be outsourced?
The current BIG guidelines (version 9, October 2025) list outsourcing as a budget head but publish no percentage cap; budget caps are checked during the site visit. The earlier version 8 guidelines (July 2020) capped outsourcing at 30% of project cost. Confirm the limit for your call before you plan.
Should a startup outsource the core experiment?
Usually not. The core experiment is where the know-how is created, and that know-how is what investors and licensees pay for. Outsource the work around it, such as analytics, sequencing or specialised characterisation, and keep the critical method and its optimisation in your own hands.
What should a contract research agreement say about IP?
That all results, data and inventions arising from the work belong to the startup, that the provider keeps them confidential, and that raw data and methods are handed over. Agree this before any material leaves your lab.